Received: from vmb.brl.mil by VMB.BRL.MIL id aa05908; 24 Jun 90 0:38 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa05545; 24 Jun 90 0:06 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa05507; 23 Jun 90 23:58 EDT Received: from [131.170.14.2] by VGR.BRL.MIL id aa04103; 23 Jun 90 23:50 EDT Received: by godzilla Date: Sun, 24 Jun 90 13:57:32 EST From: Mike Gigante Message-Id: <9006240357.16590@godzilla> To: info-iris@BRL.MIL Subject: Utah Raster Toolkit <-> SGI Image lib In a couple of following messages there are 2 programs, fromurt.c and tourt.c which I knocked up earlier this year. No guarantees etc etc. Send ny improvements back to me please! Mike Gigante, RMIT Advanced Computer Graphics Centre   Received: from vmb.brl.mil by VMB.BRL.MIL id aa05964; 24 Jun 90 0:48 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab05545; 24 Jun 90 0:06 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab05507; 23 Jun 90 23:58 EDT Received: from godzilla.cgl.rmit.oz.au by VGR.BRL.MIL id aa04110; 23 Jun 90 23:51 EDT Received: by godzilla Date: Sun, 24 Jun 90 13:58:34 EST From: Mike Gigante Message-Id: <9006240358.16597@godzilla> To: info-iris@BRL.MIL Subject: fromurt.c /* * Convert a Utah Raster Toolkit image to a SGI image file * * Mike Gigante, RMIT ACGC, Feb 1990 */ #include #include "svfb_global.h" #include "image.h" unsigned short rbuf[4096]; unsigned short gbuf[4096]; unsigned short bbuf[4096]; int x_size, y_size; /* Size of image. */ static short bg[3]; #define NRED(scan, i) (short)((sv_globals.sv_alpha) \ ? scan[0][i]*(scan[-1][i]/255.0) + \ bg[0]*(1.0 - scan[-1][i]/255.0) \ : scan[0][i]) #define NGREEN(scan, i) (short)((sv_globals.sv_alpha) \ ? scan[1][i]*(scan[-1][i]/255.0) + \ bg[1]*(1.0 - scan[-1][i]/255.0) \ : scan[1][i]) #define NBLUE(scan, i) (short)((sv_globals.sv_alpha) \ ? scan[2][i]*(scan[-1][i]/255.0) + \ bg[2]*(1.0 - scan[-1][i]/255.0) \ : scan[2][i]) main(argc,argv) int argc; char **argv; { int y; IMAGE *image; FILE *infile; unsigned char **rgb, *scan[4]; register unsigned char *src; register unsigned short *dest; short i, j, ncolors; rgb = scan+1; if(argc != 3) { fprintf(stderr,"usage fromurt infile outfile\n"); exit(1); } if ( argv[1] != NULL ) if ( (infile = fopen( argv[1], "r" )) == NULL ) { perror( argv[1] ); exit( 1 ); } sv_globals.svfb_fd = infile; if ( rle_get_setup( &sv_globals ) < 0 ) { fprintf(stderr, "fromurt: Error reading setup information from %s\n", argv[1]); exit(1); } ncolors = sv_globals.sv_ncolors > 3 ? 3 : sv_globals.sv_ncolors; bg[0] = sv_globals.sv_bg_color[0]; bg[1] = sv_globals.sv_bg_color[1]; bg[2] = sv_globals.sv_bg_color[2]; /* * Compute image size */ x_size = (sv_globals.sv_xmax - sv_globals.sv_xmin + 1); y_size = (sv_globals.sv_ymax - sv_globals.sv_ymin + 1); if ( argv[2] != NULL ) if ( (image = iopen(argv[2],"w",RLE(1),3,x_size,y_size,3)) == NULL) { perror( argv[2] ); exit( 1 ); } /* * Set up for rle_getrow. Pretend image x origin is 0. */ for (i = 0; i < 4; i++) scan[i] = (unsigned char *) malloc(x_size); sv_globals.sv_xmax -= sv_globals.sv_xmin; sv_globals.sv_xmin = 0; /* For each scan line, pack RGBs into the image memory. */ while ((y = rle_getrow(&sv_globals, scan+1)) <= sv_globals.sv_ymax) { for ( i = 0; i < x_size; i++ ) { rbuf[i] = NRED(rgb, i); gbuf[i] = NGREEN(rgb, i); bbuf[i] = NBLUE(rgb, i); } putrow(image,rbuf,y,0); /* red row */ putrow(image,gbuf,y,1); /* green row */ putrow(image,bbuf,y,2); /* blue row */ } iclose(image); }   Received: from vmb.brl.mil by VMB.BRL.MIL id aa06073; 24 Jun 90 1:03 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab05964; 24 Jun 90 0:53 EDT Received: by VMB.BRL.MIL id aa05931; 24 Jun 90 0:40 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa05799; 24 Jun 90 0:11 EDT Received: from godzilla.cgl.rmit.oz.au by VGR.BRL.MIL id aa04118; 23 Jun 90 23:55 EDT Received: by godzilla Date: Sun, 24 Jun 90 14:02:09 EST From: Mike Gigante Message-Id: <9006240402.16616@godzilla> To: info-iris@BRL.MIL Subject: acknowledgements for sgi image <-> urt I should have mentioned that some of the code is snarfed from getiris which is part of the urt distribution. The stuff included in the URT header in tourt.c was inspired by BRL-CAD's pix-rle program. Mike   Received: from vmb.brl.mil by VMB.BRL.MIL id aa06162; 24 Jun 90 1:24 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac05545; 24 Jun 90 0:07 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ac05507; 23 Jun 90 23:58 EDT Received: from godzilla.cgl.rmit.oz.au by VGR.BRL.MIL id aa04112; 23 Jun 90 23:52 EDT Received: by godzilla Date: Sun, 24 Jun 90 13:59:24 EST From: Mike Gigante Message-Id: <9006240359.16604@godzilla> To: info-iris@BRL.MIL Subject: tourt.c /* * tourt - convert from SGI Imagelib to Utah Raster Toolkit... * * Mike Gigante, RMIT ACGC, Feb 1990 */ #include "image.h" #include "svfb_global.h" short rbuf[4096]; short gbuf[4096]; short bbuf[4096]; rle_pixel pixels[4][4096]; static struct sv_globals outrle; #define outfp outrle.svfb_fd static char comment[128]; static long now; static char *who; extern char *getenv(); main(argc,argv) int argc; char **argv; { register IMAGE *image; register int y, xsize, ysize; register int z, zsize; int i; rle_pixel *rows[3]; if( argc != 3 ) { fprintf(stderr,"usage: tourt infile outfile\n"); exit(1); } if( (image=iopen(argv[1],"r")) == NULL ) { fprintf(stderr,"tourt: can't open input file %s\n",argv[1]); exit(1); } xsize = image->xsize; ysize = image->ysize; zsize = image->zsize; if(zsize<3) { fprintf(stderr,"tourt: this is not an RGB image file\n"); exit(1); } if( (outfp = fopen(argv[2],"w")) == NULL ) { fprintf(stderr,"tourt: can't open output file %s\n",argv[2]); exit(1); } outrle.sv_ncolors = 3; SV_SET_BIT(outrle, SV_RED); SV_SET_BIT(outrle, SV_GREEN); SV_SET_BIT(outrle, SV_BLUE); outrle.sv_background = 0; /* don't use background */ outrle.sv_alpha = 0; /* we don't have alpha channel */ outrle.sv_ncmap = 0; /* no color map */ outrle.sv_cmaplen = 0; outrle.sv_cmap = (rle_map *) 0; outrle.sv_xmin = outrle.sv_ymin = 0; outrle.sv_xmax = xsize - 1; outrle.sv_ymax = ysize - 1; outrle.sv_comments = (char **) 0; /* Add comments to the header file, since we have one */ now = time(0); sprintf(comment, "creation_date=%24.24s", ctime(&now)); rle_putcom(strdup(comment), &outrle); if ((who = getenv("USER")) != (char *) 0) { sprintf(comment, "created_by=%s", who); rle_putcom(strdup(comment), &outrle); } else { if ((who = getenv("LOGNAME")) != (char *) 0) { sprintf(comment, "created_by=%s", who); rle_putcom(strdup(comment), &outrle); } } sprintf(comment, "converted from SGI Imagelib file %s", argv[1]); rle_putcom(strdup(comment), &outrle); sv_setup(RUN_DISPATCH, &outrle); for (i = 0; i < 3; i++) { rows[i] = pixels[i]; } for(y=0; y Message-Id: <9006240711.16666@godzilla> To: info-iris@BRL.MIL Subject: G0 libraries As fasr as I know, these are only of use if you want to do dynamic loading (because teh MIPS linker fixes the global pointer/variable size at link time. If you attempt to subsequently load files into a running executable, there won't be enough space for the global pointers in the library. The G0 libraries solve this by being compiled with a gloal space size of 0 I could be wrong, and this could only be a partial solution. This much is true as I needed to use the G0 libraries for a large lisp application that required dynamic loading.. Mike p.s. when are the version >= 2.0 compilers from MIPS appearing on the SGI machines? Are they part of 3.3? I need 2.0's *true* alloca ...   Received: from vmb.brl.mil by VMB.BRL.MIL id aa06941; 24 Jun 90 4:19 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa06810; 24 Jun 90 3:37 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa06793; 24 Jun 90 3:27 EDT Received: from godzilla.cgl.rmit.oz.au by VGR.BRL.MIL id aa04575; 24 Jun 90 3:18 EDT Received: by godzilla Date: Sun, 24 Jun 90 17:25:23 EST From: Mike Gigante Message-Id: <9006240725.16686@godzilla> To: info-iris@BRL.MIL Subject: Panel Library from NASA Ames I see someone from SGI suggested the panel library for a customer's application in a previous message in this group. I have used the panel library extensively - having written nearly a dozen applications using it. I am very grateful to the author of the panel library (Dave Tristram) - it has been a valuable prototyping tool. However, I don't really think it is worth starting from scratch with it at this point. SGI have previously announced in this group that a SGI toolkit will be available as part of the next (post 3.3) release. This toolkit will have C++, C and f77 bindings. Pascal I'm not sure about. From the little I know of this toolkit, it seems like it should be much better than the panel library and the resultant applications will fit nicely into the SGI environment. The panel library suffers from not haveing any documentation and also being very clumsy for many things. To make things work as you would like is difficult, some common interaction styles almost impossible. The code ends up being quite ugly as your application will need to muck with the internals of the panel library a *lot* if you wish to have fine control. That is given the high overheads to start with thge panel library and that a significantly better toolkit is not more than 6 months away, I can't recommend it. However, until now it has been our only real choice given the SGI's X implementation and no alternative interaction stuff. (Sorry, I guess that the NeWS stuff has been there for a while, but I've never tried it. Wise move since it will dissappear eventually..) Mike Gigante, RMIT Advanced Computer Graphics Centre   Received: from vmb.brl.mil by VMB.BRL.MIL id aa09311; 24 Jun 90 20:24 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa09141; 24 Jun 90 20:13 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa09122; 24 Jun 90 20:01 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa06251; 24 Jun 90 19:57 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA13527; Sun, 24 Jun 90 16:52:02 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 24 Jun 90 23:01:28 GMT From: Trevor Paquette Organization: The University of Calgary Subject: night program just posted Message-Id: <1990Jun24.230128.17598@calgary.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL The night program that I just posted might not work on a 4d/20. It works fine on the 240gtx at work (64meg memory) but seems to kill the window manager on the 4d/20. Be forewarned. ___________________________________________/No man is a failure who has friends Trevor Paquette ICBM:51'03"N/114'05"W|I accept the challange, body and soul, {ubc-cs,utai,alberta}!calgary!paquette|to seek the knowledge of the ones of old paquette@cpsc.ucalgary.ca | - engraved on the Kersa Blade of Esalon   Received: from vmb.brl.mil by VMB.BRL.MIL id aa09387; 24 Jun 90 20:55 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab09141; 24 Jun 90 20:13 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab09122; 24 Jun 90 20:01 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa06253; 24 Jun 90 19:57 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA13499; Sun, 24 Jun 90 16:51:40 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 24 Jun 90 22:44:24 GMT From: Trevor Paquette Organization: The University of Calgary Subject: night program Message-Id: <1990Jun24.224424.16666@calgary.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Awhile back Howard Look posted a twilight program. I have since hacked it to bits and come up with a new version that has a 'twinkle' feature. It can also have things like 'comets' etc added to it fairly easily. If you write any enhancements to this program please email me them so I can keep track of what all the changes are. Watch out for the .signature at the end. -------------------- cut here -------------------- /* Original program from: howardl@landis.csd.sgi.com (Howard Look) Here's yet another little background program. It looks like the sky at sunset. It will only run properly on 24 or more bit plane systems. Enjoy. night enhancements uunet!alberta!calgary!paquette Jun-19-1990: Trevor Paquette - hacked up data structures for 'square stars' - added twinkle feature Please email any enhancements to me. Compile it with: cc -o night night.c -lm -lgl_s -lc_s -lmalloc */ #include #include #include #include #include #include /* a star is actually a "square" p1(x,y,r,g,b,br,bg,bb) p4(r,g,b,br,bg,bb) +----------------------------+ |<---^------ rad ----------->| | | | | rad | | | | | V | +----------------------------+ p2(r,g,b,br,bg,bb) p3(r,g,b,br,bg,bb) */ typedef struct _dummy { float x, y, rad; /* location and radius of star */ int p1br, p1bg, p1bb; /* background color behind star */ int p1r, p1g, p1b; /* color of star */ int p2br, p2bg, p2bb; int p2r, p2g, p2b; int p3br, p3bg, p3bb; int p3r, p3g, p3b; int p4br, p4bg, p4bb; int p4r, p4g, p4b; } starstruct; #define ITIMER_REAL 0 #define Y1 0.0 #define Y2 0.2 #define SMALL_MAX_BLINK 128 #define BIG_MAX_BLINK 3 struct itimerval new_value, old_value; /* event timers */ int numgroups = 10; /* num of groups of 256 stars */ int numbig = 200; /* num of big stars */ int twinkle_rate = 27; /* twinkle rate of small stars */ starstruct *small_stars = NULL; /* pointer to locations of small stars */ starstruct *big_stars = NULL; /* pointer to locations of big stars */ starstruct *small_blinker[SMALL_MAX_BLINK]; /* at most MAX_BLINK stars blink at a time */ starstruct *big_blinker[BIG_MAX_BLINK]; /* at most MAX_BLINK stars blink at a time */ main(argc, argv) int argc; char *argv[]; { int gid, skyevent(), i; short val; long dev; char *malloc(); /* set time for timeout */ new_value.it_value.tv_sec = 0; new_value.it_value.tv_usec = twinkle_rate*10000; new_value.it_interval.tv_sec = 0; new_value.it_interval.tv_usec = 0; for(i = 0; i < SMALL_MAX_BLINK; i++) { small_blinker[i] = NULL; } for(i = 0; i < BIG_MAX_BLINK; i++) { big_blinker[i] = NULL; } small_stars = (starstruct *) malloc(numgroups*256*sizeof(starstruct)); big_stars = (starstruct *) malloc(numbig*sizeof(starstruct)); imakebackground() ; /* foreground();*/ gid = winopen("") ; RGBmode(); gconfig(); shademodel(GOURAUD); ortho2(0.0, 1.0, 0.0, 1.0) ; qenter(REDRAW,gid); while (1) { dev = qread(&val) ; if(dev == REDRAW) draw_background() ; } } draw_background() { int i,j, skyevent(), col[3]; static int first = 1; static int orange[] = {255,72,0}; static int blueish[] = {0,110,189}; static int black[] = {0,0,0}; static int red[] = {255,0,0}; float v1[2],v2[2],v3[2],v4[2]; register int counter; v1[0] = v4[0] = 0.0; v2[0] = v3[0] = 1.0; v1[1] = v2[1] = 0.0; v3[1] = v4[1] = Y1; bgnpolygon(); c3i(red); v2f(v1); v2f(v2); c3i(orange); v2f(v3); v2f(v4); endpolygon(); v1[1] = v2[1] = Y2; bgnpolygon(); c3i(orange); v2f(v4); v2f(v3); c3i(blueish); v2f(v2); v2f(v1); endpolygon(); v3[1] = v4[1] = 1.0; bgnpolygon(); c3i(blueish); v2f(v1); v2f(v2); c3i(black); v2f(v3); v2f(v4); endpolygon(); if(first) setup_sky(); counter = 0; for(j=0; j= x)) found = 1; } return(ret); } skyevent() { /* anything like comets .. astral events etc,, should be added here */ long event, trand(); register int starnum, i, counter; static small_numstars = 0, big_numstars = 0; char *malloc(); int col[3]; float v1[2], v2[2], v3[2], v4[2]; signal(SIGALRM,SIG_IGN); event = trand(1, 100); /* an event is based on percentage chance of happening */ /* star twinkle = all of the time */ /* comet = 2% ?? */ /* percentages should total 100 */ /* turn on the stars that we turned off last timer we were here */ if(small_numstars) { for(i = 0; i < small_numstars; i++) { bgnpoint(); c3i(&small_blinker[i]->p1r); v2f(&small_blinker[i]->x); endpoint(); } } if(big_numstars) { for(i = 0; i < big_numstars; i++) { bgnpolygon(); v1[0] = big_blinker[i]->x; v1[1] = big_blinker[i]->y; v2[0] = big_blinker[i]->x; v2[1] = big_blinker[i]->y - big_blinker[i]->rad; v3[0] = big_blinker[i]->x + big_blinker[i]->rad; v3[1] = big_blinker[i]->y - big_blinker[i]->rad; v4[0] = big_blinker[i]->x + big_blinker[i]->rad; v4[1] = big_blinker[i]->y; c3i(&big_blinker[i]->p1r); v2f(v1); c3i(&big_blinker[i]->p2r); v2f(v2); c3i(&big_blinker[i]->p3r); v2f(v3); c3i(&big_blinker[i]->p4r); v2f(v4); endpolygon(); } } small_numstars = trand(1, SMALL_MAX_BLINK); /* blink between 1 and 256 stars */ for(i = 0; i < small_numstars; i++) { starnum = trand(0, numgroups*256); small_blinker[i] = &small_stars[starnum]; bgnpoint(); c3i(&small_stars[starnum].p1br); v2f(&small_stars[starnum].x); endpoint(); } big_numstars = trand(1, BIG_MAX_BLINK); /* blink between 1 and 256 stars */ for(i = 0; i < big_numstars; i++) { starnum = trand(0, numbig); big_blinker[i] = &big_stars[starnum]; bgnpolygon(); v1[0] = big_blinker[i]->x; v1[1] = big_blinker[i]->y; v2[0] = big_blinker[i]->x; v2[1] = big_blinker[i]->y - big_blinker[i]->rad; v3[0] = big_blinker[i]->x + big_blinker[i]->rad; v3[1] = big_blinker[i]->y - big_blinker[i]->rad; v4[0] = big_blinker[i]->x + big_blinker[i]->rad; v4[1] = big_blinker[i]->y; c3i(&big_blinker[i]->p1br); v2f(v1); c3i(&big_blinker[i]->p2br); v2f(v2); c3i(&big_blinker[i]->p3br); v2f(v3); c3i(&big_blinker[i]->p4br); v2f(v4); endpolygon(); } if((event >= 1) && (event <= 2)) { /* shooting star event */ } if((event >= 2) && (event <= 10)) { /* satilite event (see above)*/ } /* reload the timer */ signal(SIGALRM,skyevent); setitimer(ITIMER_REAL,&new_value,&old_value); } setup_sky() { register int i, j; RGBvalue tr, tg, tb; register int counter; long trand(); /* setup big stars , should be done BEFORE small stars */ for(j =0; j Message-Id: <9006251459.AA19462@tom.dallas.sgi.com> To: info-iris@tom.dallas.sgi.com Subject: VME board Hi: Has anyone out there used a 1553B VME interface board in a SGI system? Did you write or buy the driver? I'm looking for any and all input on 1553B . Thanks in advance ..... -- Tom Reed SGI - Dallas email: treed@sgidal.dallas.sgi.com vmail: 8705 phone: 214-788-4122   Received: from vmb.brl.mil by VMB.BRL.MIL id aa19279; 25 Jun 90 13:34 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa18770; 25 Jun 90 13:23 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa18704; 25 Jun 90 13:02 EDT Received: from rpi.edu by VGR.BRL.MIL id aa09365; 25 Jun 90 12:53 EDT Received: from rdrc.rpi.edu by rpi.edu (4.1/SM46-RPI-ITS); id AA20661; Mon, 25 Jun 90 12:53:19 EDT for info-iris@brl.mil Received: from ihmad.rdrc.rpi.edu by rdrc.rpi.edu (4.0/1.2-RPI-RDRC) id AA02554; Mon, 25 Jun 90 12:53:20 EDT From: Ahmad Safi Date: Mon, 25 Jun 90 12:59:28 EDT Message-Id: <2470.9006251659@ihmad.rdrc.rpi.edu> To: info-iris@BRL.MIL Subject: night program Hi: I compiled and tried to run the "night" program on a sgi 4d/50 GT. It doesn't do anything apparently. What is it supposed to do when you run it? Thanks. -- SIGNATURE : Ahmad Safi - asafi@ihmad.rdrc.rpi.edu o __ __ _______ /_/_/_/__| | | | /_ |_ | \ / |_) | \____/ ___|______| |   Received: from vmb.brl.mil by VMB.BRL.MIL id ab20064; 25 Jun 90 14:06 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac19573; 25 Jun 90 13:55 EDT Received: from adm.brl.mil by VMB.BRL.MIL id aa19571; 25 Jun 90 13:44 EDT Received: from ucbvax.Berkeley.EDU by ADM.BRL.MIL id aa06930; 25 Jun 90 13:31 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA11817; Mon, 25 Jun 90 10:24:18 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 25 Jun 90 16:58:10 GMT From: "Bernard J. Duffy" Organization: University of Maryland, Baltimore County Subject: Re: G0 libraries Message-Id: <3513@umbc3.UMBC.EDU> References: <9006240711.16666@godzilla> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9006240711.16666@godzilla> mg@GODZILLA.CGL.RMIT.OZ.AU (Mike Gigante) writes: > >As fasr as I know, these are only of use if you want to do dynamic loading >(because teh MIPS linker fixes the global pointer/variable size at link time. >If you attempt to subsequently load files into a running executable, there >won't be enough space for the global pointers in the library. The G0 libraries >solve this by being compiled with a gloal space size of 0 > >I could be wrong, and this could only be a partial solution. This much is >true as I needed to use the G0 libraries for a large lisp application >that required dynamic loading.. Does this mean programs performing dynamic loading will bomb if they weren't compiled -G 0 or would they just yield the wrong numbers/ fault results? If they bomb, I'll just leave them off the system until some user complains about their program not working due to a dynamic loading error (hopefully an informative sys-err message on failure). I currently don't have any flavors of lisp running, but might have KCL in the future (added to our machines by our KCL support person). > >Mike Thanks... Bernie Duffy Systems Programmer II | Bitnet : BERNIE@UMBC2 Academic Computing - L005e | Internet : BERNIE@UMBC2.UMBC.EDU Univ. of Maryland Baltimore County | UUCP : ...!uunet!umbc3!bernie Baltimore, MD 21228 (U.S.A.) | W: (301) 455-3231 H: (301) 744-2954   Received: from vmb.brl.mil by VMB.BRL.MIL id aa21611; 25 Jun 90 15:17 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa21230; 25 Jun 90 15:07 EDT Received: from adm.brl.mil by VMB.BRL.MIL id aa21169; 25 Jun 90 14:53 EDT Received: from ucbvax.Berkeley.EDU by ADM.BRL.MIL id aa08225; 25 Jun 90 14:20 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA14524; Mon, 25 Jun 90 11:03:32 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 25 Jun 90 17:44:19 GMT From: Paul Haeberli Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Utah RLE format Message-Id: <62804@sgi.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL This program converts UTAH RLE format imge files into IRIS image file format. . . . paul haeberli paul@sgi.com begin 777 fromutah M`6``"29;(S@````````````X``OA(#4#!`:&```("&/A(#4CX6`V(^&@-`,$`"0``````P0&E0`0"`A M````#0/@``@`````)[T`"`/@``@`(/@A`^``"```````````)[W_H*^P`"0H M@0`#`*"`(:^_`#ROM``TK[4`.*^R`"ROLP`P%"``!*^Q`"@\$1``$```"28Q M"<`\$1``)C$)P#P%$``DI0`P#^`$5`(@("$,$!I4)`0``8X$``0/X`1,)X6` M0#P!$``40``'K"(!*#P%$``DI0!8#^`$5`(@("$,$!I4)`0``3P5$``FM0#T M#!`!)`*@("$"H"`A#!`&7">E`%@00``'`````#P%$``DI0!P#^`$5`(@("$, M$!I4)`0``3P.$``\#Q``C>\!$(W.`10\$Q``C@0`"(YS`0P!SQ`C)$(``208 M``,F%@$0D!@$!#!`1;20'``,\`Q``C&,! M$#P9$`"/.0$4`$"@(0,C""H4(``J`&"0(3P0$``F$!*0CZ4`6`P0`YP"H"`A MCZ@`6`(`*"&-!```#!`!$@)@,"$"@"`A`@`H(0(@,"$,$!A,```X(8^I`%@" M`"@AC20`!`P0`1("8#`A`H`@(0(`*"$"(#`A#!`83"0'``&/J@!8`@`H(8U$ M``@,$`$2`F`P(0*`("$"`"@A`B`P(0P0&$PD!P`"/`L0`(UK`10F4@`!`7(( M*A`@_]HF,0`!#!`05`*`("$,$!I4```@(8^_`#R/L``DC[$`*(^R`"R/LP`P MC[0`-(^U`#@#X``()[T`8`#`$"$00``(),;__Y".````P!`A),;__R2$``$D MI0`"%$#_^J2N__X#X``(`````````````````^``"*^D`````````````">] M_X"OOP`X`&R.#0``CAD`!"6N__XG+``"K@P`!*X.``"'KP!L```````/P@,` M#V(`,8W_`#,9`/\#+7`EIZX`;)(8``P`````,P\`$!'@``2'K`!L$``!\R0" M__V'K`!L)`',4A&!``,`````$``![20"__^.&0```````"LA``T0(``*```` M`">D`'`D!0`!)`8`#0(`."$/X`1:KZL`@(^K`(`0```P`#:X, M``2N&```D@X`#``````QS0`0$:``!(>O`'`0``&^)`+__(>O`'```$@A``]B M`S&9`/\`#\(`AZP`M`'`S+O\` M`PYH)0`/8@,QF0#_``_"`(>L`':GK0!R,P[_``,N:"4`#'H#,?@`_P`,R@"G MK0!T,R[_`).O`'D##F@EIZT`=JUO``2-;``$`````!F``!23K0!X,3D`_P!`A,2T`!R0/``&03@`X`:]@!#&9`/\!V<`EH%@` M.(UO``0E*0`!`2\(*A0@__`Q.0#_DZT`>``````QK``"%8``-`````"3I`!Y M``!((:^I`&"OJP"`#^`$@@`$(("/JP"`CZD`8*UB``B3I`!Y```````$($(` M!"!`#^`$@B2$``&3I@!YCZD`8``&,$*/JP"``$!`(0`&,$`DQ@`!KZ@`9`!` M("$D!0`!#^`$6@(`."&3K@!YCZ@`9(^I`&"/JP"`$<``#@$`("$``!`A`0`8 M(8UX``B0>0```P)X(:WY``"3K0!Y)2D``0$M""LD0@`$%"#_]R1C``$!`"`A M#^`$@*^K`("/JP"`$```').X`'@\#!``C8P2(``````1@``'``````(`("$/ MX`4LKZL`@(^K`(`0```1D[@`>(X"````````)$+__P1!``>N`@```@`@(0_@ M!#BOJP"`CZL`@!````:3N`!XC@(`!``````D3@`!K@X`!).X`'@`````,QD` M`A,@``23KP!X$```"ZU@`!"3KP!X`````#'M``$1H``%)`X``20,``(0```# MK6P`$"0.``&M;@`0D[@`>``````S&0`$$R``!P````"1;0!7)`\``36L`("M M;P`,$````J%L`%>M8``,AZX`<`````"M;@`4A[@```TX!!```.HD`O_^C60`+```2"&OJ0!@KZL`@*^J M`$`D!0`"`4`P(0_@!%H"`#@ACZH`0(^I`&"/JP"`&4``0Y.L`'@Q10`#$*`` M$0`)&$``"1A`C6P`+"4I``$!@Q`AE$0``"1C``(`!,("``3*`#,M_P`S#@#_ M`"6D3P``C6P`+``````!@Q`AE$4``@`` M````!<("``7*`#,M_P`S#@#_`"44:/_4I$\`!I.L`'@`````,9@`"!,``(`` M````CAD````````O(0`"$"``"@`````GI``\)`4``20&``("`#@A#^`$6J^K M`("/JP"`$```#8>F`#R.#@`$`````(7-````````IZT`/(X8``".#P`$)QG_ M_B7L``*N#``$KAD``(>F`#PD`?_^``9R`P`&>@`Q[/\`,B`#BGI@`\``0D`P_@!(*OJP"`CZL`@!1```D` M0$`AAZ8`/#P$$``\!1``)*4!Z`_@!%0DA`G`$```820"__Z'I@`X``!((:^I M`&"OJ`!<`0`@(20%``$"`#@A#^`$6J^K`("'I@`\CZ@`7(^I`&`!!A@ACZL` M@`$#""L0(``*`0`0(9!8````````%P```@`````E*0`!)$(``0!#""L4(/_X M`````"4I``&OJ0!@KZ@`7``)((`/X`2"KZL`@(^K`("/J`!0`P)00``:\H M``"'K@`\`(`0(0$.&"$`@P@K$"``$"0)``&03___`````!7@``@`````C6T` M,``)8(`!K,`AKP(``(>Y`#PE*0`!`1D8(21"``$`0P@K%"#_\@````"-;@`P M``EX@`'/:"$0```"K:```*U@`#"-;```!AKX"`L)(9``P`````,RX`$!7```0D#P`!$```!``` M$"$D#P`!H6\`8"0"__R/OP`)(0)P`_@!%0`8#`A$```+X^_`!0\!!`` M/`40`"2E`GDDA`G`#^`$5`!@,"$0```GC[\`%#P$$``\!1``)*4"IB2$"<`/ MX`14`&`P(1```!^/OP`4/`00`#P%$``DI0*_)(0)P`_@!%0`8#`A$```%X^_ M`!0\!!``/`40`"2E`O,DA`G`#^`$5`!@,"$0```/C[\`%(^N`!@`````)<\` M!"WA``40(/_S```````/>(`\`1```"\((8PO`````````>``"`````"/OP`4 MCZ(`&`/@``@GO0`8)[W_Z*^_`!04P``"KZ4`'">&@&`,$`$DKZ8`((^F`""/ MI0`<#!`#2`!`("$00``#`$`@(0P0&E0`````C[\`%">]`!@#X``(``````/@ M``BOA("P)[W_:*^V`"@`@+`AK[\`+*^U`"2OL@`@KZ4`G([2`#2.U0`4KZ`` MA([.`!`D`0`"%<$`-`````"OL0!(CL\`#``````1X``,`````)+8`%<````` M,QD`@!,@``<`````CZ@`G([%`!B-!/_\```P(0P0!E(DI0`!CL(`!```B"$8 M0``>`````#(I`/\%(0`"`2`((20A``<``5##`LI8(9%L`#@R+0`')`X``0&N M>`0!C\`D$P``#0````"/N0"``!P````".PP!8`````"1C``&NPP!8$``"6`!@ M$"&OL`!$K[$`2*^S`$"OM``\K[<`.*^^`#22V``X`````#,9``$3(``'``"8 M(8^H`)P`````C0D``!````(!-9@A``"8(3P4$``FE`G`CDH````````I00`" M$"``"``````GI`!X)`4``20&``(/X`1:`D`X(1````V22``,CDX`!">K`'B5 MP0```````*5A``".6```CDP`!"<9__XECP`"KD\`!*Y9``"22``,`````#$) M`!`1(`(.AZ8`>"0*``$0``(>HLH`8##-`$`1H``EA[D`>(Y+````````+6$` M`A`@``@`````)Z0`?"0%``$D!@`"#^`$6@)`."$0```-AZD`?(Y.``0````` MA``` M```S*`#_``A6```*3@,Q*P#_KLL`7(^-@+``````$:``"8>B`'B.Q@!"0!_[\``A(#$``!W`!!$"2'I@!X M)`$`_S#&`/\`!C8```8V`S#&`/\4P0`#KZ8`A"0,__^OK`"$CZ\`A"0+``$Q M^`#_!P$``@,`""$D(0`'``'(PP+90"&1"@`X,>D`!P$K:`2.U0`4`4UP)!'` M``<`````CZP`G``/P(`!F,@ACR@````````!%9@ACXN`L``````18`&Z)`$` M`3P%$`"/I@"$)*4#-@*`("$/X`14`J`X(8>B`'@D`?^_``(2`Q```:X`01`D M,,D`0!$@`#6'I@!XCDH````````M00`"$"``"``````GI`!\)`4``20&``(/ MX`1:`D`X(1````V'J`!\CDT`!`````"%K@```````*>N`'R.6```CD\`!"<9 M__XE[``"KDP`!*Y9``"'J`!\```````(6@,`"%(`,4W_`#%I`/\!+7`ECXN` ML``.?```#L0```]D`P`8S`.GK@!\`JRH(1%@``@">9@A/`40```.-```!C0# M)*4#5P*`("$/X`14`J`X(8>B`'@D`?^_``(2`Q```7@`01`DAZ8`>(^(@+`P MQ@#_``8V```&-@,PQ@#_`J:H(1$``6\"9I@A/`40`"2E`V\"@"`A#^`$5`*@ M."&'H@!X)`'_OP`"$@,0``%E`$$0)(^J`(0D#``!,4D`_P4A``(!(`@A)"$` M!P`!:,,PV`!`,4\`!P'L\`03```C`LVX(8Y9````````+R$``A`@``@````` M)Z0`?"0%``$D!@`"#^`$6@)`."$0```-AZP`?(Y+``0`````A6X```````"G MK@!\CDT``(Y(``0EJO_^)0D``JY)``2N2@``AZP`?```````#'H#``S*`#,K M_P`Q^`#_`PMP)0`.C`"GK@!\$```!P`1C`.'L0!X`````#(Q`/\`$8X``!&. M`S(Q`/^2Z``X)C$``0$>2"01(``@``````)@("$D!0`!`B`P(0_@!%H"0#@A M,BT``1&@`$@`````/`H0`(U*$B``````$4``!0`````/X`4L`D`@(1```#\` M````CD(````````D0O__!$$`!:Y"```/X`0X`D`@(1```#8`````CD(`!``` M```D3P`!$```,:Y/``22S`!A`````!&```TF(@`!)B4``02A``(`H`@A)"$` M`0`!*$,`!2A``D`@(0_@!%XD!@`!$```(@`````F(@`!!$$``@!`""$D(0`! M``$00P`"$$`80``:`$"`(3P9$`"/.1(@`````!,@``4`````#^`%+`)`("$0 M```/)A#__XY"````````)$+__P1!``6N0@``#^`$.`)`("$0```&)A#__XY" M``0`````)%@``:Y8``0F$/__'@#_Z`````"/BX"P`G&8(1%@`#P"L:@ADNX` M.``````!WD`D$0``,0`````\!1```G&`(R2E`X<"@"`A`B`P(0_@!%0"H#@A M&B``"0````"2!@```H`@(2>%@&8/X`14)A```28Q__\>(/_Y`````#P)$`"- M*1(@`````!$@``8`````)`0`"@_@!2X"@"@A$```'8>B`'@\`A``C$()P#P! M$``D0O__!$$`!JPB"<`D!``*#^`$0@*`*"$0```2AZ(`>#P*$`"-2@G$)`T` M"CP/$`"A30``C>\)Q#P!$``E[``!$```!ZPL"<0\!1``)*4#G@*`("$"(#`A M#^`$5`*@."&'H@!X)`'_OP`"$@,0``"9`$$0)(^Y`(0D"``!,S@`_P?_^)9@``JY8``2N60``AZ@`?```````"'(# M``AJ`#&J_P`QR0#_`2IX)0`/C`"GKP!\$```!P`1C`.'L0!X`````#(Q`/\` M$8X``!&.`S(Q`/^.3````K&H(2V!``(0(``()K4``2>D`(`D!0`!)`8``@_@ M!%H"0#@A$```#8>I`(".6``$`````(<+````````IZL`@(Y(``".60`$)0W_ M_BI`("/F8"P``E2`P`)8@`QF/\`,4\`_P'X6"43(``* MIZL`@``+=```#D0#/`40`"2E`[6OJ``0`H`@(28F``$/X`14`J`X(9+M`#@` M`````;Y0)!%``!$J(0`*%"``"`````"'I@"``F`@(0P0!E(F)0`!`G&8(1`` M``@FB`'B'J0"`)C'__R9S``$&(?_\HFG__X>B`'@D`?^_``(2 M`Q```"(`01`D)`P``:+,`&"'H@!X)`'_OP`"$@,0```;`$$0)#P%$``DI0/7 M#^`$5`*`("$,$!I4)`0``8>B`'@D`?^_``(2`Q```!``01`DAZ8`>"0!_[\` M!C(#`,$0)"1/__\MX0`'$"#_[@``````#WB`/`$0```O""&,+P`4``````'@ M``@`````)`$``1!!``,D`0`'%$']Q0````".P@!8C[``1(^Q`$B/LP!`C[0` M/(^W`#B/O@`T`````(^_`"R/L@`@C[4`)(^V`"@#X``()[T`F`"@$"$80``& M)*7__P"@$"$DI?__H(8``!Q`__PDA``!`^``"``````GO?_8K[\`%*^E`"P` M@#`AC,<`&(S(``P``!@A$0``!R3G``&0S@!7`````#'/`(`1X``"`````"0# M``&,Q0`$```0(1B@`!,`````,%@`_P] M`"@#X``(`````">]_^BOOP`4C((`#`"@,"$00``"`````"3&__R,C@`$```8 M(0'"*"$8H``/``````#`$"&,1````````!"```]_W"OL@`4`("0(:^_`!ROLP`8CE,`-">(@,PGKP!\)`[,4J>. M@,ROCX#$KXB`R)$"``$!X"`AHZ(`?)$#```D!0`")`8``0)@."$/X`1FHZ,` M?8Y"`!`D`0`"%$$`!``````D`@`!$```!Z.B`(@40``$```0(1````(D`@`" M```0(:.B`(B.6``,`````!,```8`````DZ(`B``````T0@`$,$(`_Z.B`(B. M0@`P`````!!```H`````C%D````````3(``&`````).B`(@`````-$(`"#!" M`/^CH@"(CDD`!"0*``BCJ@"*HZD`B8Y+`"0`````&6``"P````".3``L```` M`!6```<`````/`00`#P%$``DI00`#^`$5"2$"<"N0``DCDT`*">(@,RCK0", MCDX`)">D`("CK@"+CD\`%"0%``VGCX#,D0(``20&``&CH@"`D0,```)@."&C MHP"!CE@`'`````"GF(#,D0(``2>X`(:CH@""D0,```````"CHP"#CDD`%(Y9 M`!@``````RE0(R5+``&GBX#,D0(``0````"CH@"$D0,```````"CHP"%CDT` M'(Y,`""OB(#(`8UP(R7/``&GCX#,KYB`Q)$"``$`````HZ(`AI$#```/X`1F MHZ,`AXY9`!``````$R``)0````"OL``LCD0`!`_@!((DA``!CD@`!`!`@"$9 M```.```H(0``&"$`0"`ACDD`""2E``$!(U`AC4L``"1C``2@BP``CD@`!"2$ M``$`J`@J%"#_]@`````%`0`"`0`((20A``$``2A#``4H0"2E``$"`"`A)`8` M`0_@!&8"8#@A#^`$@`(`("&/L``L$```'8Y"`"0\#!``C8P2(``````1@``& M````````("$/X`4N`F`H(1```!..0@`DCF(````````D0O__!$$`!JYB```` M`"`A#^`$0@)@*"$0```)CD(`)(YM``0`````H:```(YN``0`````)<\``:YO M``2.0@`D`````!A``&4`````K[``+*^Q`"B.6``H)!D``0,92`0!(@`9``"` M$@`0($`/X`2"`````!1```D`0(@A/`00`#P%$``DI00U)(0)P`_@!%0"`#`A M#!`:5"0$``$:``!%```H(3(&``,0P``3``400``%$$`"(A@ACDH`+"2E``$! M0E@AE60``"1"``(`!&(",(X`_P`.>@`QC0#_`:_():1Y```4Q?_T)&,``A"P M`#("("`A``400`(B&"$`$$!`CE@`+"1C``@#`D@AE20`````````!%(",(P` M_P`,<@`Q2P#_`6YH):1M__B.3P`L``````'BR"&7)0`"```````%P@(PJ@#_ M``IB`#,)`/\!+%@EI&O_^HY.`"P``````<)H(96F``0```````9Z`C#8`/\` M&%(`,?D`_P,J2"6D:?_\CDP`+``````!@E@AE6<`!B1"``@`!W(",.\`_P`/ MP@`QS0#_`;C()11(_]2D>?_^`B`@(0(`*"$D!@`"#^`$9@)@."$/X`2``B`@ M(8^P`"R/L0`H`````).B`(@`````,$H`"!%``%&/OP`I`$2OB8#$KXB`R*>1@,R1`@`!`2`@(:.B`$21`P``)`4``B0& M``$"8#@A#^`$9J.C`$6.4``P`````(X$````````$(``#3(L``$/X`3"```` M`(X$```D!0`!)$8``0_@!&8"8#@AC@0`!"80``04@/_V`````#(L``$1@``= MC[``+#P+$`"-:Q(@`````!%@``8````````@(0_@!2X"8"@A$```$X^P`"R. M8@```````"1"__\$00`&KF(`````("$/X`1"`F`H(1````F/L``LCFX`!``` M``"AP```CF\`!``````E[0`!KFT`!(^P`"R/L0`H`````(^_`!R/L@`4C[,` M&`/@``@GO0"0)[W_X*^D`""/K@`@K[\`'*^P`!B,L``T*<$!`!0@`'`````` M/`\0`(WO$B``````$>``!@`````D!`!!#^`%+@(`*"$0```2`````(X"```` M````)$+__P1!``:N`@``)`0`00_@!$("`"@A$```"`````".&0`$)!@`0:,X M``"."``$`````"4)``&N"0`$/`H0`(U*$B``````$4``!@```````"`A#^`% M+@(`*"$0```2`````(X"````````)$+__P1!``:N`@`````@(0_@!$("`"@A M$```"`````"."P`$`````*%@``".#``$`````"6-``&N#0`$/`\0`(WO$B"/ MK@`@$>``!J>.@."3A(#A#^`%+@(`*"$0```3`````(X"````````)$+__P1! M``:N`@``DX2`X0_@!$("`"@A$```"0````"3@H#ACA@`!`````"C`@``CAD` M!``````G*``!K@@`!#P)$`"-*1(@`````!$@``8`````DX2`X`_@!2X"`"@A M$```3(^_`!R.`@```````"1"__\$00`&K@(``).$@.`/X`1"`@`H(1```$*/ MOP`]`"`GO?_@/`X0`(W.$B"OOP`4C*,`-!'```H`@#`AKZ,`'"0$``(`8"@A M#^`%+J^F`""/HP```$)`\``J,/ M``",>0`$`````"]`"`# MX``(`````">]_^"OOP```!@`````D!`!##^`% M+@(`*"$0```2`````(X"````````)$+__P1!``:N`@``)`0`0P_@!$("`"@A M$```"`````".&0`$)!@`0Z,X``"."``$`````"4)``&N"0`$/`H0`(U*$B`` M````$4``!@```````"`A#^`%+@(`*"$0```2`````(X"````````)$+__P1! M``:N`@`````@(0_@!$("`"@A$```"`````"."P`$`````*%@``".#``$```` M`"6-``&N#0`$/`\0`(WO$B"/K@`@$>``!J>.@."3A(#A#^`%+@(`*"$0```3 M`````(X"````````)$+__P1!``:N`@``DX2`X0_@!$("`"@A$```"0````"3 M@H#ACA@`!`````"C`@``CAD`!``````G*``!K@@`!#P)$`"-*1(@`````!$@ M``8`````DX2`X`_@!2X"`"@A$```3(^_`!R.`@```````"1"__\$00`&K@(` M`).$@.`/X`1"`@`H(1```$*/OP`]`"`GO?_@K[\`%(RC`#00@``\C[\`%#P. M$`"-SA(@`````!'```@`````)`0``0!@*"$/X`4NKZ,`'(^C`!P0```4```` M`(QB````````)$+__P1!``BL8@``)`0``0!@*"$/X`1"KZ,`'(^C`!P0```( M`````(QX``0D#P`!HP\``(QY``0`````)R@``:QH``0\"1``C2D2(``````1 M(``&`````"0$``$/X`4N`&`H(1```!./OP`4C&(````````D0O__!$$`!JQB M```D!``!#^`$0@!@*"$0```)C[\`%(QK``0D"@`!H6H``(QL``0`````)8T` M`:QM``2/OP`4)[T`(`/@``@`````)[W_X*^E`"2/K@`DK[\`'*^P`!BOI``@ MC-``-!'``+R/OP`8 M@."3A(#A#^`%+@(`*"$0```3`````(X"````````)$+__P1!``:N`@``DX2` MX0_@!$("`"@A$```"0````"3@H#AC@@`!`````"A`@``C@D`!``````E*@`! MK@H`!#P+$`"-:Q(@`````!%@``8`````DX2`X`_@!2X"`"@A$```4(^E`"2. M`@```````"1"__\$00`&K@(``).$@.`/X`1"`@`H(1```$:/I0`DDX*`X(X, M``0`````H8(``(X-``0`````):X``1```#RN#@`$/`\0`(WO$B``````$>`` M!@`````D!``%#^`%+@(`*"$0```2`````(X"````````)$+__P1!``:N`@`` M)`0`!0_@!$("`"@A$```"`````".&0`$)!@`!:,X``"."``$`````"4)``&N M"0`$/`H0`(U*$B``````$4``!P````"3I``G`@`H(0_@!2XDA/__$```%X^E M`"2.`@```````"1"__\$00`(K@(``).D`"<"`"@A)(3__P_@!$(PA`#_$``` M"X^E`"23H@`GC@L`!"1"__\P0@#_H6(``(X,``0`````)8T``:X-``2/I0`D MCZ0`("2E``$$H0`"`*`((20A``$``2A#)`8``@_@!&8"`#@AC[\`'(^P`!@# MX``()[T`(">]_^"OI0`DCZX`)*^_`!ROL``8KZ0`(*^F`"B,\``T*<$!`10@ M`*D`````/`\0`(WO$B``````$>``!@`````D!`!&#^`%+@(`*"$0```2```` M`(X"````````)$+__P1!``:N`@``)`0`1@_@!$("`"@A$```"`````".&0`$ M)!@`1J,X``"."``$`````"4)``&N"0`$/`H0`(U*$B``````$4``!@`````` M`"`A#^`%+@(`*"$0```3CZX`)(X"````````)$+__P1!``:N`@`````@(0_@ M!$("`"@A$```"8^N`"2."P`$`````*%@``".#``$`````"6-``&N#0`$CZX` M)#P8$`"/&!(@)<___Q,```:GCX#@DX2`X0_@!2X"`"@A$```$P````".`@`` M`````"1"__\$00`&K@(``).$@.$/X`1"`@`H(1````D`````DX*`X8X9``0` M````HR(``(X(``0`````)0D``:X)``0\"A``C4H2(``````10``&`````).$ M@.`/X`4N`@`H(1```!,`````C@(````````D0O__!$$`!JX"``"3A(#@#^`$ M0@(`*"$0```)`````)."@."."P`$`````*%B``".#``$`````"6-``&N#0`$ M/`\0`(WO$B"/K@`@$>``!J>.@."3A(#A#^`%+@(`*"$0```3`````(X"```` M````)$+__P1!``:N`@``DX2`X0_@!$("`"@A$```"0````"3@H#ACA@`!``` M``"C`@``CAD`!``````G*``!K@@`!#P)$`"-*1(@`````!$@``8`````DX2` MX`_@!2X"`"@A$```B8^_`!R.`@```````"1"__\$00`&K@(``).$@.`/X`1" M`@`H(1```'^/OP`,@."3A(#A#^`% M+@(`*"$0```3`````(X"````````)$+__P1!``:N`@``DX2`X0_@!$("`"@A M$```"0````"3@H#AC@X`!`````"AP@``C@\`!``````E^``!KA@`!#P9$`"/ M.1(@`````!,@``8`````DX2`X`_@!2X"`"@A$```%(^_`!R.`@```````"1" M__\$00`&K@(``).$@.`/X`1"`@`H(1````J/OP`LH@``)`0`!P_@!$*OI0`C#P``C+D`!``````G*``!K*@`!#P)$`"-*1(@`````!$@``4`````#^`% M+@``("$0```2C[\`%(RB````````)$+__P1!``6LH@``#^`$0@``("$0```) MC[\`%(RJ``0`````H4```(RK``0`````)6P``:RL``2/OP`4)[T`(`/@``@` M`````````">]_GBOI`&(CZX!B*^^`""OOP`DKZ4!C`#`0"$5P``'``#P(8T/ M`%B/N`&,``````'XR"$0``&YK1D`6*^R`$2-"0!<`````!4@`!D`````CZ0! MC"0!``,`@0`:KZ@!D```(!(`!"!```0@0`_@!((DA``$CZ!`(292``$"0P@J%"#_YR80``&/L`!`$```"H^L`7R-#@!<)`L``:7```"/ MN0&,C0H`7"#X M"0````"/IP&0`````*S@`%B/J`&0`````(T.``R-`P`$$<```P`````0```# M)`3__Z^@`70``"`A`(,(*A`@`10`````K[8`+*^P`$"OL0`\K[,`.*^T`#2' MM@!DK[4`,*^W`"@PF0#_`[E((9$I`'0`````$2``^0````"/J@&(``18@`%+ M8"&-EP``KZ@!D*^D`72OH`&`C0T``#P/$```#<"``PW`(0`8P,`!^'@AC>\` MV`$`*"$!X/@)`````(^G`9``````C.X`7`````"%Q````````!B```Z/JP%\ MC/D``#P*$```&4B``3E((0`)2,`!25`AC4H`W```*"$!0/@)```P(8^G`9`` M````CZL!?``````98`"[CZT!="5L__^OK`!(KZ``6(SM`%R/N`!8``"@(0&X M$"&$50``A%,``@*@D"$"=0@J%"``6R0!__\0``!/)`'__P*R""H0(``'`O*` M(9(/````````%?X``P`````0``!,)!0``9(>````$K0`$```2``6M`,"\H`A MD@(```/`B"$440`>`````(SN```\"1````[(@`,NR"$`&"&-[P#H``"@(0/`("$"5B@C`>#X"0``,"&/ MIP&0DAX````2M```$JP``!:T`Q````@`%:P#)`'__Q*!_^<"\H`A$H#_KR0! M``$2@?^[`O*`(292``$"<@@J$"#_]B0!__\6@0`/`````(SN```\"1````[( M@`,NR"$`&(`!^'@A``]XP`'/<"&-S@#<)H8``2S&``$D1/__`<#X"20%``&/ MIP&0`````(^Y`8"/J@!8CZP!?"] M_\BOL@`<`*"0(:^S`""OL0`8`,"((0"`F"$:0``PK[\`)*^T`"P\%!``)I0` MS*^P`#".+@!@)!@"`#'/`?\##X`C`E`(*A`@``X`````CB<`-`)@("$"0"@A M#^`$9B0&``&..0!@``````,R0"&N*`!@C[0`+(^P`#`0```9C[\`)(XG`#0" M8"`A`@`H(0_@!&8D!@`!CBL``(XI`&``"V"``8M@(0`,8,`!,%`AKBH`8`*, M:"&-K@`@`G"8(0'`^`D"("`A`E"0(QY`_]D`````C[``,(^T`"P`````C[\` M)(^Q`!B/L@`,J@``/`P0```*6(`!:E@A``M8P`&+ M8"&-C`#0`*`@(0&`^`D`````C[\`%">]`!@#X``(``````/@``BOI```)[W_ MZ*^_`!2,C@``/!@0```.>(`![G@A``]XP`,/P"&/&`#P``````,`^`D````` MC[\`%">]`!@#X``(``````"`0"$8X``7```0(8^C`!``````D&X````````1 MP``-```````"((``I,@ACRD```#$>"$!*%`AD4L``(WX````````$PL`!"1" M``$#X``(```0(21"``$41__N)&,``:^C`!`D`@`!`^``"```````@%@A`*!` M(0#`4"$$X`!E```0(8^D`!``````%$``!``"<(`0```&```8(0`"<(``CG@A MA>/__@`````D8P`!%$<`!``"P(`0```&)0;__P`"P(``F,@AAR8````````D MQO__`,,(*A0@``H`8$@A`6E@(9&-````````%4T`!@#)""HE*0`!`,D(*A`@ M__DEC``!`,D(*A0@`#X`````)&\``0'I""H``G"`%"```@".*"$40``0`$<( M*A`@``L`X!@A``/`@`"88"&%F?_\A8W__B1C__\`0P@JI9D``"6,__P4(/_Y MI8T`!J2I```0```#).<``21"__\DI?_\`,D(*A0@``LE+___`6E@(9&.```` M````$4X`!B4O__\E*0`!`,D(*A`@__DEC``!)2___P$F""H4(``6I*\``B3F M__\`1@@J$"``$@````"$N``&)/G__@!9""HD0P`"$"``"Z2X``(``VB``(U@ M(86.``"%CP`")&,``0!G""JECO_\)8P`!!0@__FEC__Z`,`X(21"``$`X@@J M$"#_GP`````#X``(`.`0(8^C`!``@$`A`*!((0#`4"$`X%@A&&``^@``$"$P M;0`#%:``"0````#'C8`0QXR`%,>#@!C'@H``@"00```^`0(H(<>! M@"#'@(`DQX.`&,>"@!S'C8`0QXR`%`$"*"$!(C`A`4(X(0%B8"&0K@`````` M`$2.(``%P0`%1H`AH<>)@"C'B(`L`````$8H,8"0SP``1B`R@D2/@``%X0`% M1H"$H<>%@"C'A(`L`````$8DE(!&(I("D/@```````!$F(``1BA1@`2@"P`````1C(A`$8L(H(D0@`!)*4``23&``$DYP`!1BHR`"6, M``%$6?@`1%GX```````W(0`#."$``D3!^```````1B!$)$0$@`!$V?@`%:+_ MT:&$__\00P"S``````$"*"$!(C`A`4(X(0%B8"&0K@```````$2.D``%P0`% M1H"1(<>'@"C'AH`L`````$8F(0"0SP``1B`B@D2/0``%X0`%1H!$(<>3@"C' MDH`L`````$8RA`!&(H&"D/@```````!$F$``1B91``0 M@"P`````1C"4@$8LDH)&*B&`1%GX`$19^```````-R$``S@A``)$P?@````` M`$8@,B1$!$``1-GX`*&$``"0K@`!`````$2.@``%P0`%1H"$H<>%@"C'A(`L M`````$8DE("0SP`!1B"2@D2/,``%X0`%1H`R(<>1@"C'D(`L`````$8P0@!& M(D$"D/@``0````!$F#``1B14@`(@"P`````1BB$`$8L M@H)&*I$`1%GX`$19^```````-R$``S@A``)$P?@``````$8@(:1$!#``1-GX M`*&$``&0K@`"`````$2.0``%P0`%1H!$(<>3@"C'DH`L`````$8RA`"0SP`" M1B""@D2/(``%X0`%1H`AH<>)@"C'B(`L`````$8H,8!&(C2"D/@``@````!$ MF"``1C)4```*,>&@"P`````1B9"`$8L0H)&*H2`1%GX`$19 M^```````-R$``S@A``)$P?@``````$8@D21$!"``1-GX`*&$``*0K@`#```` M`$2.,``%P0`%1H`R(<>1@"C'D(`L`````$8P0@"0SP`#1B!"@D2/D``%X0`% M1H"1(<>'@"C'AH`L`````$8F(0!&(B0"D/@``P````!$F)``1C!2``$@"P`````1B0Q@$8L,H(D0@`$)*4`!"3&``0DYP`$1BI$`"6, M``1$6?@`1%GX```````W(0`#."$``D3!^```````1B"$I$0$D`!$V?@`%$/_ M4Z&$__\#X``(`````````````````````">]_^"OL``0`("`(:^_`!0,$!"X M`@`@(0(`("$,$!0U```H(98.`'``````,<\``A'@`#<`````AA@`<@`````3 M```$`@`@(0P0$SP"`"`A`@`@(0(`*"$,$!/Y)`8`F(89`'(`````$R```P`` M```,$!,\`@`@(98(``(D`0$`,0G_`!4A`"(``````@`@(0P0%#4D!0(`E@H` M")8+``H``````4L`&0``8!(`#&B`KZT`'(8.`'(`````$<``!`````".!`"0 M#!`3(P&@*"&.!0"0CZ8`'`P0$_D"`"`AA@\`<@`````1X``%`````(X$`)2/ MI0`<#!`3(P````".!0"4CZ8`'`P0$_D"`"`AC@0`@``````0@``#``````_@ M!(``````C@0`A``````0@``#``````_@!(``````EA@``B0!`0`S&?\`%R$` M!P````".!`"0#^`$@`````".!`"4#^`$@`````".!`!L#!`:8``````/X`2` M`@`@(8^_`!2/L``0`^``"">]`"`GO?_HK[\`%)2.`'``````,<\``A'@`!B/ MOP`4C(4`@``````0H``4C[\`%(R8`'P`````$P``$(^_`!2$A@!VA(<`>`P0 M&$ROI``8CZ0`&`````"4F0`&`````!!9``:/OP`4E(@`<``````U"0`@I(D` M<(^_`!0GO0`8`^``"``````GO?_HK[\`%)2.`'``````,<\``17@``,````` M$```,R0"__^,F`"``````!<```8`````#!`2_*^D`!B/I``8`````*R"`("$ MA@!VE)D`"```````V0@K%"``$@````"$@@!XE(@`"B1"``$``A0```(4`P!( M""ND@@!X%"``!Z2``':4B0!PI(``>#4J`!"DB@!P$```%20"__^$A@!V```` M`(R%`("$AP!X#!`9;J^D`!B/I``8`````)2+``:,@P"`A(T`=B5L__^L@P!\ M):X``:2.`':DC`!ZE&(``"1O``*LCP!\,$+__X^_`!0GO0`8`^``"``````G MO?_HK[\`%*^E`!R4C@!P`````#'/``(5X``#`````!```#PD`O__C(4`@``` M```4H``(``````P0$OROI``8CZ0`&`````"L@@"`C(4`@`````",F`!\```` M`!,``"``````A(8`=H2'`'@,$!A,KZ0`&(^D`!@`````A((`=I29``@D0@`! M``(4```"%`,`60@K%"``$*2"`':$@@!XE(@`"B1"``$``A0```(4`P!(""ND M@@!X%"``!Z2``':4B0!PI(``>#4J`!"DB@!P$```$"0"__^,A0"``````)2+ M``:LA0!\)6S__Z2,`'J/K0`<`````*2M``",C@!\`````"7/``*LCP!\CZ(` M'`````"/OP`4)[T`&`/@``@`````)[W_Z*^_`!0`@!@A)&0`&`_@!,@D!@!0 MC[\`%">]`!@#X``(``````/@``BLA0!H)[W_V*^E`"ROI@`P`.`8(8^N`#B/ MKP`\C[@`0*^D`"B/I0`HCZ<`,(^F`"ROOP`D```@(:^C`!"OK@`4KZ\`&`P0 M$9>ON``]`"@#X``(`````">]_]BOI@`P`.`8(8^N`#B/KP`\C[@` M0*^E`"R/I@`LCZ<`,*^_`"0``"@AKZ,`$*^N`!2OKP`8#!`1EZ^X`!R/OP`D M)[T`*`/@``@`````)[W_V*^_`!2OI``H)`0`F*^P`!"OI0`LKZ8`,`_@!(*O MIP`T`$"`(0!`("$/X`00)`4`F(^N`#``````D<\``0`````Y^``K+Q@``1,` M``BON``@/`00`#P%$``DI02`#^`$5"2$"<`,$!I4)`0``8^Y`#`D`0!WDR@` M```````5`0!=CZL`+(^F`"P`````$,``%(^J`"@`P"`A#!`::"0%`;:/J0`@ MKZ(`*!$@``H`````!$``"`````"/I``H#!`:8`````"/I``L#!`:<"0%``*O MH@`HCZ8`+`````"/J@`H``````5!``F/H@`X/`00`#P%$``DI02H#^`$5"2$ M"<`,$!I4)`0``8^B`#@D"P':I@L``(^L`#0D`P`!I@P``H^M`#PL00`"I@,` M"J8#``@4(``$I@T`!H^N`$``````I@X`""Q!``,4(``$`````(^O`$0````` MI@\`"I88``H`````%P,`""0)``.6"``()!D``A4#``6F&0`$$````Z8#``0D M"0`#I@D`!#P*`)@U2I:`K@H`#*X``!`"`"`A#!`182>%@'"N```4I@``"0&`)@D`0"8$$$`58^K`"`\!!``/`40`"2E!,P/X`14)(0) MP`P0&E0D!``!$```3(^K`""/JP`L`````!%@``V/K0`HCZP`(``````1@``$ M```H(1````(D!0`"```H(8^D`"P,$!IP`````*^B`"B/K0`H``````6A``2/ MI``H$```T```$"&/I``H`@`H(0P0&H`D!@"8)`$`F!!!``@`````/`00`#P% M$``DI03T#^`$5"2$"<`,$!I4)`0``98&```D`0':,,\`_P`/P@``!G("`=C( M)1$` M!208``$D#@`"$````Z8.`'`D&``!IA@`<)89``(D`0$`,RC_`!4!`',````` ME@D`")8*``H``````2H`&0``$!(``A"``$`@(0_@!(*OH@`JX``'RN``"`#!`2_`(`("&/I``H)`@"`*X(`(BF``!TI@`` M=J8``'BN`@"$)`4"````,"$,$!J(K@0`;`(`$"&/OP`4C[``$`/@``@GO0`H M)[W_X*^_`!0`@!@AE&(`!@```````G&"`<(@(0_@!((`!""`%$``"*^B`!P\ M!!``/`40`"2E!:@/X`14)(0)P`P0&E0D!``!C[\`%(^B`!P#X``()[T`(``% M$$,`@#`A&$``#```&"&4Q```)&,``0`#'````QP#``1R`@`$>@`!S\`E`&(( M*J38```4(/_V),8``@/@``@```````40@P"`,"$80``4```8(3P'`/\``W"` M`,XH(8RD```D8P`!``3"`C,9_P``!'X"``1*``$G4"0!^4`E``,<```#'`,! M"E@E``1F``%L:"4`8@@J%"#_[ZRM```#X``(`````">]_^BOI``8CZ0`&*^_ M`!0,$!,2)`4`#(^D`!@D!0`,#!`3(R2$``R/I``8)`4`!`P0$R,DA`!HC[\` M%">]`!@#X``(`````">]_^"OOP`4KZ4`)">%@'@,$!%MKZ8`*!1```BOH@`< M/`00`#P%$``DI07,#^`$5"2$"<`,$!I4)`0``8^N`!R/N``DE<\`!@````"O M#P``C[D`'(^I`"B7*``(`````*TH``"/OP`4)[T`(`/@``@`````)[W_Z*^_ M`!2$@@!Z`````"1"__\``A0```(4`P1```>D@@!ZC(,`?`````"48@``)&X` M`A````.LC@!\#!`0V0````"/OP`4)[T`&`/@``@`````)[W_Z*^_`!2$@P!Z M`````"1C__\``QP```,<`P1@``FD@P!ZC(X`?#"C__^EPP``C(\`?`!@$"$E M^``"$````ZR8`'P,$!$8`````(^_`!0GO0`8`^``"``````GO?_HK[\`%`"@ M."$`X"@AKZ<`'*^D`!@,$!/7KZ8`((^D`!B/I@`@CZ<`')2#``*D@`!T,&+_ M`*2&`'@40``6I(<`=I2"``:4CP`(`,(`&3!I`/\``'`2```````````!SP`9 M``#`$@```````````.(`&0``R!(#.$`A``````$)`!D``"@2#!`4-22E`@`0 M```8C[\`%"0!`0`400`-`````)2+``B,B@"0`,L`&0``8!(!AV@A``UP@`%. M>"&-Y0``#!`4-0`````0```)C[\`%#P$$``\!1``)*4%]`_@!%0DA`G`#!`: M5"0$``&/OP`4)[T`&`/@``@`````)[W_Z*^_`!0`@!@AE&X`"`"@0"$!#@@K M$"``!@#`."&4;P`*``````#O""L4(``3C[\`%#P$$``\!1``)*4&$"2$"<`! M`#`A#^`$5*^C`!B/HP`8/`00`#P%$`"49@`(E&<`"B2E!C@/X`14)(0)P`P0 M&E0D!``!C[\`%">]`!@#X``(`````">]_^"OI``@CZX`(*^F`"BOOP`4CZ8` M*(W$`&P,$!IX`````(^O`"BOH@`<$$\`"8^C`"`\!!``/`40`"2E!DP/X`14 M)(0)P`P0&E0D!``!CZ,`((^Y`"B,>`"(``````,90"&L:`"(C[\`%(^B`!P# MX``()[T`(">]_^"OI``@CZX`(*^F`"BOOP`4CZ8`*(W$`&P,$!J``````(^O M`"BOH@`<$$\`"8^C`"`\!!``/`40`"2E!F@/X`14)(0)P`P0&E0D!``!CZ,` M((^Y`"B,>`"(``````,90"&L:`"(C[\`%(^B`!P#X``()[T`(">]_^BOOP`4 M`(`8(8QN`(@`````$<4`$H^_`!2,9`!LK&4`B*^E`!P,$!J(```P(01!``F/ MI0`] M`!@#X``(`*`0(2>]_^"OOP`4E(\`"(2.`'B$AP!V`<\`&8R)`)0``,`2`P?( M(0`90(`!*%`AC48````````$P0`,C[\`%#P$$``\!1``)*4&F"2$"<`/X`14 MKZ8`'`P0&E0D!``!CZ8`'`````"/OP`4)[T`(`/@``@`P!`A)[W_Z*^F`""O MI0`OI``8CZ0`&(^N`"24CP`(C[D`(`'/ M`!F,B@"0C(@`E(R)`(R/IP`<)`'__P``P!(#&2@A``40@`%"6"$!`A@AK6D` M`(QF````````$,$`!0````",C``4``````&&:"&LC0`4K&<``(R.`(P````` M`<=X(:R/`(R/OP`4)[T`&`/@``@`````)[W_Z`"@&"$D`@`!K[\`%!1B`(<` MP$@A%.(`A0````"/K@`H`(`X(0".0"$`B`@K$"``>P$@*"$`X"`A).<``@#H M""L0(``7`````)#B__^0[__^`````!7B``4`````D/@````````06``.```` M`"3G``$`Z`@K$"``"@````"0XO__D/G__@`````7(O_X`````)#J```````` M%$K_]``````DY__^`.0P(Q#``#8`````*,$`?Q0@``0`!AP`$````R0#`'X` M!AP```,<`S1K`(`H80`)H*L```##,",4(``9)*4``9",```D8__XH*P``)"- M``$``QP`H*T``9".``(``QP#H*X``I"/``,H80`)H*\``Y"8``0DI0`(H+C_ M_)"9``4DA``(H+G__9"*__X`````H*K__I"+__\0(/_IH*O__P!@$"$D8___ M``,<`!!```H``QP#`&`0(9",```D8___``,<```#'`,DA``!)*4``11`__B@ MK/__%,#_S2C!`'\`X"`A).<``0#H""N0X___$"``#@#D,".0[0```&`0(16B M``H`Y#`C).<``0#H""L0(``&`.0P(Y#N````````$<+_^0``````Y#`C$,`` M$`#H""L`8!`A*,$`?Q0@``0`!AP`$````R0#`'X`!AP```,<`P##,".@HP`` M)*4``22E``$4P/_TH*+__P#H""L4(/^(`.`@(22E``$`J1`C$``!OJ"@__\0 M8@`$)`4``A```(PD!0`")`4``A3E`(D`````CZ\`*`"`."$`CT`A`(@(*Q`@ M`'L!("@A`.`@(23G``(`Z`@K$"``%P````"0XO__D/C__@`````7`@`%```` M`)#Y````````$%D`#@`````DYP`!`.@(*Q`@``H`````D.+__Y#J__X````` M%4+_^`````"0ZP```````!1+__0`````).?__@#D,",0P``V`````"C!`'\4 M(``$``8<`!````,D`P!^``8<```#'`,T;`"`*&$`":2L````PS`C%"``&22E M``*0C0``)&/_^*2M``"0C@`!``,<`*2N``*0CP`"``,<`Z2O``20F``#*&$` M":2X``:0F0`$)*4`$*2Y__B0B@`%)(0`"*2J__J0B__^`````*2K__R0C/__ M$"#_Z:2L__X`8!`A)&/__P`#'``00``*``,<`P!@$"&0C0``)&/__P`#'``` M`QP#)(0``22E``(40/_XI*W__A3`_\THP0!_`.`@(23G``$`Z`@KD./__Q`@ M``X`Y#`CD.X```!@$"$5P@`*`.0P(R3G``$`Z`@K$"``!@#D,".0[P`````` M`!'B__D``````.0P(Q#``!``Z`@K`&`0(2C!`'\4(``$``8<`!````,D`P!^ M``8<```#'`,`PS`CI*,``"2E``(DI0`"%,#_]*2B__X`Z`@K%"#_B`#@("$D MI0`"`*D0(P1!``(`0`@A)"$``0`!$$,0``$OI*#__A1E`)``````%.(`C@`` M``"/N``H`(`X(0`8R$``F4`A`(@(*Q`@`(,!("@A`.`@(23G``0`Z`@K$"`` M%P````"4XO_^E.K__``````50@`%`````)3K````````$$L`#@`````DYP`" M`.@(*Q`@``H`````E.+__I3L__P`````%8+_^`````"4[0```````!1-__0` M````).?__`#D,",$P0`"`,`((20A``$``3!#$,``-@`````HP0!_%"``!``& M'``0```#)`,`?@`&'````QP#-&X`@"AA``F@K@```,,P(Q0@`!DDI0`!E(\` M`"1C__B@KP``E)@``@`#'`"@N``!E)D`!``#'`.@N0`"E(H`!BAA``F@J@`# ME(L`""2E``B@J__\E(P`"B2$`!"@K/_]E(W__`````"@K?_^E([__A`@_^F@ MKO__`&`0(21C__\``QP`$$``"@`#'`,`8!`AE(\``"1C__\``QP```,<`R2$ M``(DI0`!%$#_^*"O__\4P/_-*,$`?P#@("$DYP`"`.@(*X3C__X0(``.`.0P M(Y3X````8!`A%P(`"@#D,",DYP`"`.@(*Q`@``8`Y#`CE/D````````3(O_Y M``````#D,",$P0`"`,`((20A``$``3!#$,``$`#H""L`8!`A*,$`?Q0@``0` M!AP`$````R0#`'X`!AP```,<`P##,".@HP``)*4``22E``$4P/_TH*+__P#H M""L4(/^``.`@(22E``$`J1`C$```GJ"@__\490"4`````!3E`)(`````CZH` M*`"`."$`"EA``(M`(0"(""L0(`"#`2`H(0#@("$DYP`$`.@(*Q`@`!<````` ME.+__I3L__P`````%8(`!0````"4[0```````!!-``X`````).<``@#H""L0 M(``*`````)3B__Z4[O_\`````!7"__@`````E.\````````43__T`````"3G M__P`Y#`C!,$``@#`""$D(0`!``$P0Q#``#8`````*,$`?Q0@``0`!AP`$``` M`R0#`'X`!AP```,<`S1X`(`H80`)I+@```##,",4(``9)*4``I29```D8__X MI+D``)2*``(``QP`I*H``I2+``0``QP#I*L`!)2,``8H80`)I*P`!I2-``@D MI0`0I*W_^)2.``HDA``0I*[_^I2/__P`````I*___)28__X0(/_II+C__@!@ M$"$D8___``,<`!!```H``QP#`&`0(929```D8___``,<```#'`,DA``")*4` M`A1`__BDN?_^%,#_S2C!`'\`X"`A).<``@#H""N$X__^$"``#@#D,".4Z@`` M`&`0(15"``H`Y#`C).<``@#H""L0(``&`.0P(Y3K````````$6+_^0`````` MY#`C!,$``@#`""$D(0`!``$P0Q#``!``Z`@K`&`0(2C!`'\4(``$``8<`!`` M``,D`P!^``8<```#'`,`PS`CI*,``"2E``(DI0`"%,#_]*2B__X`Z`@K%"#_ M@`#@("$DI0`"`*D0(P1!``(`0`@A)"$``0`!$$,0```)I*#__CP$$``\!1`` M)*4&N"2$"<`/X`14`&`P(0P0&E0D!``!C[\`%">]`!@#X``(`````">]_^@` MH!@A)`(``:^_`!0`@$`A%&(`5`#`2"$4X@!2``````$`,"$!(!`AD,0``"3& M``$PA0!_,*7__Q"@`56/OP`4,(X`@!'``"H`````+*$`"!0@`!H`H!@AD,\` M`"2E__B@3P``D-@``3"E__^@6``!D-D``BRA``B@60`"D,H``R1"``B@2O_[ MD,L`!"3&``B@2__\D,S__0````"@3/_]D,W__@````"@3?_^D,[__Q`@_^F@ M3O__`*`8(22E__\08/_9,*7__P"@&"&0SP``)*7__S"E__\DQ@`!)$(``11@ M__F@3___$`#_T)#$``"0Q```+*$`"!0@``XDQ@`!)*7_^#"E__\LH0`(H$0` M`*!$``&@1``"H$0``Z!$``2@1``%H$0`!J!$``<0(/_T)$(`"`"@&"$DI?__ M$&#_NC"E__\`H!@A)*7__S"E__^@1```%&#_^R1"``$0`/^SD,0``!!B``0D M!``"$```520$``(D!``"%.0`4@`````!`#`A`2`0(9#$```DQ@`!,(4`?S"E M__\0H`#]C[\`%#"8`(`3```J`````"RA``@4(``:`*`8(9#9```DI?_XI%D` M`)#*``$PI?__I$H``I#+``(LH0`(I$L`!)#,``,D0@`0I$S_]I#-``0DQ@`( MI$W_^)#.__T`````I$[_^I#/__X`````I$___)#8__\0(/_II%C__@"@&"$D MI?__$&#_V3"E__\`H!@AD-D``"2E__\PI?__),8``21"``(48/_YI%G__A`` M_]"0Q```D,0``"RA``@4(``.),8``22E__@PI?__+*$`"*1$``"D1``"I$0` M!*1$``:D1``(I$0`"J1$``RD1``.$"#_]"1"`!``H!@A)*7__Q!@_[HPI?__ M`*`8(22E__\PI?__I$0``!1@__LD0@`"$`#_LY#$```49`!4`````!3B`%(` M`````0`P(0$@$"&4Q```),8``C"%`'\PI?__$*``J(^_`!0PB@"`$4``*@`` M```LH0`(%"``&@"@&"&4RP``)*7_^*!+``"4S``",*7__Z!,``&4S0`$+*$` M"*!-``*4S@`&)$(`"*!.__N4SP`(),8`$*!/__R4V/_Z`````*!8__V4V?_\ M`````*!9__Z4RO_^$"#_Z:!*__\`H!@A)*7__Q!@_]DPI?__`*`8(93+```D MI?__,*7__R3&``(D0@`!%&#_^:!+__\0`/_0E,0``)3$```LH0`(%"``#B3& M``(DI?_X,*7__RRA``B@1```H$0``:!$``*@1``#H$0`!*!$``6@1``&H$0` M!Q`@__0D0@`(`*`8(22E__\08/^Z,*7__P"@&"$DI?__,*7__Z!$```48/_[ M)$(``1``_[.4Q```%&0`5``````4Y`!2``````$`,"$!(!`AE,0``"3&``(P MA0!_,*7__Q"@`%./OP`4,(P`@!&``"H`````+*$`"!0@`!H`H!@AE,T``"2E M__BD30``E,X``C"E__^D3@`"E,\`!"RA``BD3P`$E-@`!B1"`!"D6/_VE-D` M""3&`!"D6?_XE,K_^@````"D2O_ZE,O__`````"D2__\E,S__A`@_^FD3/_^ M`*`8(22E__\08/_9,*7__P"@&"&4S0``)*7__S"E__\DQ@`")$(``A1@__FD M3?_^$`#_T)3$``"4Q```+*$`"!0@``XDQ@`")*7_^#"E__\LH0`(I$0``*1$ M``*D1``$I$0`!J1$``BD1``*I$0`#*1$``X0(/_T)$(`$`"@&"$DI?__$&#_ MNC"E__\`H!@A)*7__S"E__^D1```%&#_^R1"``(0`/^SE,0``#P$$``\!1`` M)*4&V"2$"<`/X`14`&`P(0P0&E0D!``!C[\`%">]`!@#X``(`````">]_\BO ML``8`("`(:^_`!R6#@!P`*!0(3'/`((`P&`A%>```P#@6"$0``$3)`+__Y8" M``0`````+$$``Q`@``,L00`"``!8(2Q!``(0(``"````````8"&6`P`"```` M`#!B_P`40`!U)`$!`!```&HP8@#_E@0`!HX(``R."0`0C@,`A`"`$"$!0"@A M$$``$B2$__^4N```)*4``J!X``"08@````````$B""L0(``#`$@(*P!`2"$` M2`@K$"```@``````0$`A`(`0(22$__\40/_P)&,``:X(``RN"0`0`@`@(0&` M*"$,$!.7`6`P(98"``:.!0"$`@`@(0P0$_D`0#`AE@(`!A```-R/OP`C`!@3 M(``$`````(^D`"P,$!,2`&`H(98"``80``!8C[\`%#P$$``\!1``)*4'.`_@ M!%0DA`G`#!`:5"0$``$0``!/C[\`%"0!``$0H?_'`````"0!``(0H?_?```` M`!``__``````)`$!`!1!`#P`````$```,C!E`/\,$!10`@`@(8X%`(0``C0` M``8T`PP0%!<"`"`AC@0`A(^F`"PD!0`!#!`6Y"0'``*6`@`&$```,X^_`!0, M$!10`@`@(0`"'`".!0"$``,<`P`"-```!C0#IZ,`&`P0%!<"`"`AA@@`C M`!@1```$`````(X$`(0,$!,2`&`H(8X$`(2/I@`L)`4``@P0%N0D!P`"E@(` M!A```!J/OP`4/`00`#P%$``DI0=,#^`$5"2$"<`,$!I4)`0``1```!&/OP`4 M)`$``1"A_\T`````)`$``A"A_]D`````$`#_\``````\!!``/`40`"2E!V`/ MX`14)(0)P`P0&E0D!``!C[\`%(^P`!`#X``()[T`*``````GO?U8`(`8(20. M__^OK@`<%&``$:^_`!0\!!``#^`$;"2$"8000``%`$`8(9!/````````%>`` M"`````"3F("``````!,```,`````$```(P``$"$G@X"$/`40`"2E"9"C@("` M)Z0"*`_@!+ROHP*HCZ,"J">D`C8/X`2\`&`H(2>D`B@,$!IP```H(01``!$` M0"`A)Z4`)B0&`@(,$!J`KZ0`("0!`@(400`(CZ0`(#P$$``DA`>`)Z4`)@_@ M!(PD!@("KZ``'(^D`"`,$!I@`````(^B`!P`````C[\`%">]`J@#X``(```` M```````GO?_HK[\`%`_@!#ZOI``8CZ0`&`P0&I``````C[\`%">]`!@#X``( M```````````D`@/N````#!#@``,`````"!`:E``````#X``(```0(20"`_`` M```,$.```P`````($!J4``````/@``@`````)`(#[0````P0X``#``````@0 M&I0``````^``"``````D`@/L````#!#@``,`````"!`:E``````#X``(```` M`"0"`^L````,$.```P`````($!J4``````/@``@`````)`(#^P````P0X``# M``````@0&I0``````^``"``````D`@/I````#`/@``@`````/`$0`*PB$EP# MX``()`+__SP#$`",8Q&$)`(#^0"#("$````,%.``!0`````\`1``K"01A`/@ M``@`8!`A"!`:E``````\`A``C$(1@```````@@@K$"```@``````0"`A)`(# M^0````P4X/_T`````#P!$`"L)!&$`^``"```$"$````````````````````` M/`(0`"1",I`\`0_`K"(`0#P"$``D0E*8/`$/P*PB`$0\`A``)$)RH#P!#\"L M(@!(/`(0`"1"":`\`0_`K"(`3#P"$``D0A(0/`$/P*PB`%`\`A``)$(/X#P! M#\"L(@!4/`(`0"1":F`\`0_`K"(`%#P"`$`D0FJ4/`$/P*PB`!@\`A``)$(2 M4#P!#\"L(@`0/`(0`"1"$EP\`0_`K"(`'#P"#X`D0A((/`$/P*PB`"`\`@^` M)$(2$#P!#\"L(@`D/`(/@"1"$@`\`0_`K"(`*#P"$``D0@>`/`$/P*PB```# MX``(```````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M``````````!`#9@`0`UX`$`-6`!`#3@`0`X,`$`0G`!`$9``0!)(`$`8E`!` M$VP`0!:<`$`8>'5S86=E.B!FF4@)60J)60@:6X@2!F:6QE"@`E"!T M;R`E9"D*`%-K:7`@)60@<&EX96QS("AT;R`E9"D*`%-K:7`@)60@<&EX96QS M("AT;R`E9"D*`%!I>&5L(&1A=&$@)60@*'1O("5D*3H`4&EX96P@9&%T82`E M9"`H=&\@)60I"@!2=6X@;&5N9W1H("5D("AT;R`E9"DL(&1A=&$@)3`R>`H` MF5D(&]P8V]D93H@)60*``````!#;VQO M`H` M`&EO<&5N.B!E7!E"@!I;6=L:6(Z(')O=R!N=6UB M97(@;W5T(&]F(')A;F=E("5D("5D"@``:6UA9V4@'!A;F0Z(&)A9"!B<'`Z("5D("5D"@!P=71R;W7!E"@```&=E=')O=SH@=VEE7I;7%U>7V!!0D-$149'2$E*2TQ- M3D]045)35%565UA96GM\?7Y_```````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M`````````````````$-(4D-,05-3`````"]L:6(O8VAR8VQA Organization: NASA Ames Research Center, Mtn Vw CA 94035 Subject: Wanted: location of patch database Message-Id: <52461@ames.arc.nasa.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Is there a database of patch files available? For instance, one of the SGI demos uses "bridge.dat", which appears to refers to the Golden Gate Bridge in bicubic Bezier patches. Please reply directly. Thanks, ====================================================================== Alex Woo, MS 227-2 | woo@ames.arc.nasa.gov NASA Ames Research Center | NASAMAIL ACWOO Moffett Field, CA 94035 | {seismo,topaz,lll-crg,ucbvax}! Phone: (415) 604-6010 | ames!pioneer!woo ====================================================================== {hplabs,hao,att,decwrl,allegra,tektronix,menlo70}!ames!pioneer!woo ======================================================================   Received: from vmb.brl.mil by VMB.BRL.MIL id aa01649; 25 Jun 90 22:37 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ac01515; 25 Jun 90 22:26 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01483; 25 Jun 90 22:10 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa11422; 25 Jun 90 22:04 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA17595; Mon, 25 Jun 90 18:57:44 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 01:09:15 GMT From: Trevor Paquette Organization: The University of Calgary Subject: Irix 3.3 Changes Part 1/2 Message-Id: <1990Jun26.010915.24643@calgary.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL ------------------------- Irix 3.3 Release Notes 5. Changes Taken from the command 'pcat /usr/relnotes/standard/ch5.z' and edited to remove underlining etc.. ------------------------- 5. Changes This chapter describes changes to o IRIX o networking o documentation o graphics o demos o program development tools o software installation tools o 4Dgifts o IRIS WorkSpace o 4Sight 5.1 Changes_to_IRIX A number of changes to the file system buffer cache and virtual memory have eliminated many operating system tuneable parameters, and removed the necessity of changing others. Please consult the System Configuration and Tuning Guide for more details. The 4D1-3.3 release includes these changes to IRIX: o The header files in /usr/include/bsd and its subdirectories have been moved to /usr/include. If you had old source code that incorrectly included BSD headers files with #include or you are upgrading from a previous IRIX release, install the dev.sw.bsdhdrs subsystem. This subsystem uses symbolic links to provide compatibility. It will not be supported in the next major release. o rcsdiff(1) now understands the new diff(1) -biwt options. The -q (quiet) option returns diff's exit status. o Several additions have been made to the BSD Mail program (mail_bsd(1)): - lines starting with # are ignored - new ``unread'' command - better handling of RFC822-style addresses, and ability to add all 822 header fields via tilde escape commands - support for ``reply-to'' header field - uses the TIOCGWINSZ ioctl call to determine the actual screen size - fixed mailfile locking to be compatible with the System V (mail) file locking scheme - added read-only mailfile locking to prevent multiple instances of Mail from having simultaneous write access to the mailfile - applied flock(3)'ing to all types of lock files so that the lock scheme works over NFS o System V mail (mail_att(1)) now supports ``reply-to'' and uses flock(3) for locking files, like Mail. o You can change savecore(1M)'s default location for storing kernel crash files with /etc/config/savecore.options. o The system logging daemon, syslogd(1M), can now filter log messages through a user-specified program. o All messages printed by the kernel are now routed to the logical device /dev/klog, which is read by syslogd(1M). Kernel messages are filtered through /usr/adm/klogpp which translates disk device names into mounted file system names. o syslogd(1M) has been updated to the 4.3BSD version. It now supports the routing of messages to a remote syslogd. o syslogd(1M) no longer creates the file /etc/syslog.pid containing the process id of the syslogd process. The killall(1M) command provides the ability to send a signal to a process by name, so keeping the process id of syslogd around is no longer necessary. Note that this requires changing the entry in the crontab for root which switches to a new system log file every Sunday morning. In Release 4D1-3.2, this crontab entry included the following kill(1) command to send a SIGHUP to syslogd(1M) to cause it to restart with the new log file: kill -1 `cat /etc/syslog*.pid` In Release 4D1-3.3, the corresponding portion of the crontab entry reads: killall 1 syslogd Release 4D1-3.3 contains a new version of the file /usr/spool/cron/crontabs/root that incorporates the above change, but this file is considered a configuration file. This means that when Release 4D1-3.3 is installed on a system, the file will not be replaced by the new version unless the pre- existing version has not been modified in any way. If you have made any changes to the above file to customize it for your particular needs, then after the installation of Release 4D1-3.3, the new version of the file will exist as /usr/spool/cron/crontabs/root.N. In that case, incorporate the above change so that syslogd will receive the SIGHUP that causes it to restart with the new log file. o The finger(1) command has been enhanced to be like the 4.3BSD version. The Berkeley-dependent phone number format has been removed. finger also understands both shorter and longer phone numbers. o man(1) now supports the BSD options -k and -f for performing keyword searches in the NAME sections of manual pages. The alternate forms for man -k and man -f, apropos(1) and whatis(1), are also supported. The NAME section database required to support the manual page keyword searches is created using the new command makewhatis(1M). o The default paging program used by man(1) has been changed to ul | more -s -f -k This change permits you to scan back through previously viewed pages of a man page using the scroll bar in a wsh(1) window. This change also prevents the viewing of manual pages from clobbering the log of recent commands displayed in a wsh(1) window. The old behavior can be achieved by setting the environment variable PAGER or MANPAGER to ul | page -s -f -k o As a result of the widespread recognition of the security problems associated with setuid and setgid shell scripts, support for that feature has been disabled in Release 4D1-3.3 of IRIX. An attempt to execute a shell script with the setuid (or setgid) permission bit set will result in the error EPERM, unless it happens that the owner of the file (or group of the file if setgid) happens to be equal to the effective uid (gid, respectively) of the executing processexecuting process. In Release 4D1-3.3, it is possible to re-enable the support for setuid and setgid shell scripts by reconfiguring the kernel, but this configuration mechanism is provided only to ease the transition and will likely be removed in the next major software release. If you want to allow setuid shell scripts on Release 4D1-3.3 IRIX, edit the file /usr/sysgen/master.d/kernel and change the value for the variable nosuidshells to zero. The relevant line is shipped as: int nosuidshells = 1; meaning that setuid shell scripts are not allowed by the system as it is shipped. After changing the value to be zero, rebuild the kernel using lboot(1M). o The sar(1) -t option display headers have been changed to be more mnemonic. An -h option has been added to display kernel heap statistics. o The gr_osview(1) bbuf display has been removed -- buffer cache information is now a part of the rmem display. An rmemc display has been added which is a superset of rmem. It distinguishes free memory which cache file system data from those that don't. net and netif bars display network activity. o The su(1M) command now looks at the user's .rhosts file before prompting for a password. See su(1M) for details. 5.2 Changes_to_Networking The 4D1-3.3 software release includes these changes to networking: o Security holes in rlogind(1M), rshd(1M), rcp(1M), rdist(1C), and sendmail(1M), have been fixed. o tftpd(1M) has been made more secure. It logs remote requests if the -l option is specified in /usr/etc/inetd.conf. The -s option rejects requests for files specified by an absolute pathname (the default). Also, you can restrict requests to certain directories by specifying the list on tftpd's command line in inetd.conf. See tftpd for details. o The timed(1M) daemon now automatically adjusts the frequency of the system clock for far more accurate timekeeping. o The network startup script /etc/init.d/network now starts named(1M) before mounting NFS file systems. o To specify command-line options for the inetd and rwhod daemons, create inetd.options and rwhod.options files in /etc/config. The network startup script now looks for these files when it starts the daemons. See network(1M), inetd(1M), and rwhod(1M) for details. o inetd(1M) now supports RFC1078's TCP port multiplexer, TCPMUX. It allows you to add services started from inetd using an assigned port number. The maximum number of arguments for a server is now 11 (the first argument must be the program name). See the man page for details. See ~4Dgifts/examples/network for sample TCPMUX client and server programs. o ftp(1C), ftpd(1M), rlogind(1M), and rshd(1M) are now based on the BSD Networking Release. o The file /etc/issue, if it exists on your system, is displayed at the start of rlogin(1C), telnet(1C), and ftp(1C) sessions. o Remote logins via rlogin(1C) and telnet(1C) are disabled if the file /etc/nologin exists. The file can contain a message, which will be printed before the session is terminated. Remote shell rsh(1C) accesses are also disabled but the file's contents are not displayed. o login(1) can now log remote login attempts to syslogd(1M). See the login(1) manual entry for details. o ifconfig(1M) now prints the netmask with a leading 0x, which is the same format expected by the netmask command line parameter. o telnet(1C) now negotiates the following options with the remote telnet server at startup: window size, terminal speed, local flow control, and line mode. telnetd(1M) supports these new options except for line mode. o ftp(1C) has the following new commands: chmod, idle, modtime, newer, umask, and site. o ftp(1M) now supports ``restricted'' accounts. A restricted account has access restricted to files in its home directory - it is basically an ``anonymous'' account with a password. o rshd(1M) now requires the account to have a local directory. The -n option disables keep-alives. The -l option disables use of every user's .rhosts file. o Sample versions of named(1M) database files are now installed in /usr/etc/named.d/Examples. Files in /usr/etc/named.d are not touched by the installation program (except for the README file.) Put your versions of the database files in this directory. The root.cache file has been updated - you should update your copy of it. o Various stream management bugs in named(1M) have been fixed. o The nslookup(1C) utility has been enhanced. If you enter an Internet address when the query type is set to ``A'' (address), nslookup will find its corresponding host name. Before, you had to use the PTR query type to convert the address into a name. Internally, nslookup uses PTR queries instead of now-obsolete inverse queries for the address-to-name lookup. The ls -t type command lists data records of the specified type for a domain. The default root server has been changed to ns.nic.ddn.mil. o The /usr/etc/resolv.conf file lets you specify the ordering of the Yellow Pages, BIND and file-based host-address lookup services used by gethostbyname(3N) and gethostbyaddr(3N). Also, you can now specify the default domain search list for BIND queries. See the description of hostresorder and search in the resolver(4) manual page. o gethostbyname(3N) has been enhanced to look for host aliases when doing YP, BIND or file-based lookups. In previous releases, only BIND lookups allowed aliases. See hostname(5) for details on how to specify aliases. Also, gethostbyname(3N) now returns a valid hostent structure for lookup request that are Internet addresses. This means you can specify a host's Internet address to any network command that expects a host name, such as rcp(1C) and rlogin(1C). o The Cornell super-routing daemon, gated(1M), is now supported. It is in the eoe1.sw.ipgate subsystem. gated handles the HELLO and EGP routing protocols in addition to RIP. o IRIX now has full support for IP multicasting and the IGMP protocol. This implementation is based on the Stanford Multicast 1.2 release. Note: This RFC-1112 level 2 implementation of IP multicasting/IGMP is experimental and subject to change in order to track future BSD UNIX releases. - Chapter 3 in the Network Communications Guide describes the programming interface. See ~4Dgifts/examples/network for a sample multicasting program. - A daemon for routing multicast packets, mrouted(1M), is included in the eoe1.sw.ipgate subsystem. This subsystem is needed only on machines with more than 1 network interface. - The dog(6D) demo now uses IP/UDP multicasting by default. rwhod(1M) can be configured to use multicasting instead of broadcasting. - See Chapter 7 of these release notes for known problems with multicasting. o The ping(1M) command has been enhanced. You can specify the time-to-live and output interface for pings sent to an IP multicast address. The -p option lets you specify the data to be sent. In the previous release, the optional packet size, count and preload arguments were specified after the hostname. Now they must be specified using the -s, -c, and -l options. o fingerd(1M) has a new -S option (secure) to disallow requests to see who is currently logged in. o UUCP support has been upgraded to System V Release 3.2. Additional modems, such as the Telebit T2500, have been added to the Dialers file. 5.3 Changes_to_Documentation o The TCP/IP User's Guide has been completely revised for this release and has been renamed to Network Communications Guide. It now documents UNIX domain sockets, IP multicasting, and Sun Remote Procedure Call (RPC) programming. 5.3.1 On-line_Release_Notes The options to the relnotes command for viewing on-line release notes have changed slightly. When you install the on-line documentation for a product, you can view the release notes on your screen as you would an on-line manual page. However, unlike the on- line manual pages, the printed hard copy of these release notes is more up to date than the on-line version. The relnotes command now accepts the following argments: -h describes how to use relnotes product displays a table of contents for a given product product chapter displays the given chapter of product -t product chapter sends a copy of the given chapter to the default printer To see a description of how to use the on-line release notes tool, type: relnotes -h To see which products have on-line release notes installed, type: relnotes To see which chapters of a product are installed, enter the command below, replacing the word product with a product name generated when you type the previous command. relnotes product To view a specific chapter of product, replace the words product and chapter in the following command: relnotes product chapter To page through a chapter, press and to quit, press or . See the relnotes(1) manual entry for more information. 5.4 Changes_to_Graphics This section contains information on changes to the Graphics Library and the Distributed Graphics Library. 5.4.1 Changes_to_the_Graphics_Library o All the extern routine declarations in are now prototyped. If your code calls Graphics Library routines with the wrong number of arguments, it will no longer compile. Calls to setcursor are the most likely offenders, since it is specified as ignoring its final two arguments. If you give cc the -prototypes flag as we recommend, you will get warnings when the type of an actual pointer argument to a routine is different from that specified in its prototype. e.g. long * vs. unsigned long *. Casting the argument to the correct type will eliminate the warning. Note that the form of the cast for arguments that are 2-dimensional arrays is (element-type (*)[n]), e.g. the cast for the parray argument to the poly routine is (Coord (*)[3]). o A section for obsolete symbols has been added to , , and . Symbols found in these sections should not be used in new development. In particular, you should convert all uses of the typedef Cursor to unsigned short[16]; this will prevent conflicts with X11's typedef of the same name. o The 256 vertex limit for points drawn with bgnpoint, v, endpoint and lines drawn with bgnline, v, endline bgnclosedline, v, endclosedline has been lifted. The limitation still exists for polygons. o The name space of GL devices has been partitioned as follows: 0x0000 -> 0x0FFF Devices defined by SGI 0x0001 -> 0x00FF Buttons 0x0100 -> 0x01FF Valuators 0x0200 -> 0x02FF Pseudo devices 0x0300 -> 0x0EFF Reserved 0x0F00 -> 0x0FFF Additional buttons 0x1000 -> 0x7FFF Devices defined by users 0x1000 -> 0x2FFF Buttons 0x3000 -> 0x3FFF Valuators 0x4000 -> 0x7FFF Pseudo devices 0x8000 -> 0xFFFF Cannot be used The definitions of the *COUNT symbols in have been changed so that the ISBUTTON() and ISVALUATOR() macros evaluate to TRUE for arguments that fall within the ranges specified above for buttons and valuators. o The types of a number of GL routine arguments have changed to provide a consistent interface where all color, pixels, bitmasks and device arguments are unsigned. The changes are: ________________________________________________________ Routine | Argument | Old Type | New Type ________________________________________________________ czclear cval long unsigned long defpattern mask short * unsigned short * defrasterfont raster short * unsigned short * getdev devs short * Device * linesmooth mode long unsigned long lrectread parray long unsigned long lrectwrite parray long unsigned long pagecolor pcolor short Colorindex pntsmooth mode long unsigned long qenter qtype short Device setdblights mask long unsigned long setpup mode long unsigned long zdraw b long Boolean zwritemask mask long unsigned long ________________________________________________________ o A large number of new getgdesc inquiries have been defined. see getgdesc(3G) for more information. o The z-range mapping controlled by the GLC_ZRANGEMAP mode to glcompat now occurs when a display-list is interpreted. In the previous release, it occurred when the display-list was built. o The symbol CPOSX_INVALID is no longer defined in . o Calling a GL routine on a machine that does not implement it is now always a nop; error messages are now never issued. o greset no longer partially resets the input subsystem. However, it is still reset by ginit and gbegin. See greset(3G). o gbegin now does what its manual page specifies: a ginit except for altering the color map. o mmode Upon transitioning to MVIEWING or MPROJECTION mode, the ModelView and Projection matricies are both set to the identity matrix. Upon transitioning back to MSINGLE mode, the single matrix is also set to the identity matrix. In both cases, the matrix stack is reset. o The following headers are no longer part of /usr/include/gl on the IRIS-4D B and G: file.h ic.h obj.h object.h winreq.h 5.4.2 Changes_to_the_Distributed_Graphics_Library The dglopen call now looks at the environment variable DGLBUFSIZE to obtain a default communications buffer size. If this variable is not defined, a default value of 4K is used. This call is useful over networks whose transmission unit is greater than the 4K default size. The variable DGLBUFSIZE can be specified as a decimal number indicating the number of bytes, as a decimal number immediately followed by a K or k (indicating kilobytes), or as a decimal number immediately followed by an M or m (indicating megabytes). For example, to specify a 2 megabyte buffer, DGLBUFSIZE can be set to 2097152, 2048K, or 2M. 5.5 Changes_to_Demos o dog(6D) now sends multicast packets by default instead of broadcast packets. Multicasting eliminates the load on other machines not using dog. For dogfights against 3000s and previous IRIS-4D Series releases, use the -b option to specify broadcast mode. If the system cannot set up the socket in multicast mode, it uses broadcast mode. The -t option specifies the maximum number of times the packet can be forwarded between networks. o buttonfly(6D) buttons have been reorganized and the button files have been moved out of the /usr/demos directory into a directory structure that mirrors the button hierarchy. 5.6 Changes_to_Program_Development_Tools This section explains changes to the following program development tools: o C header files o C compiler o error messages o dbx o other tools 5.6.1 Changes_to_C_Headers Nearly all the files in /usr/include have been changed to use ANSI C-style function prototypes. Since the /usr/include headers have been updated to return void * where appropriate (according to ANSI), some practices that used to work no longer work: #include main() { char *cp; int n; cp = memcpy(x,y,5) + n; } now generates an error, since memcpy returns a void * and you cannot increment a void *. cp = ((char *)memcpy(x,y,5)) + n; is (normally) the appropriate change. 5.6.2 Changes_to_the_Compilation_Environment The default space allocation for basic blocks in the optimizer was increased from 500 to 1000. Use -Olimit to alter the limit. The maximum length of a string that the compilers will emit in assembler code (see the -S flag to cc(1)) has been increased from 256 bytes to 1024 bytes. If your source code results in a line length in assembler that is larger than your chosen viewer can handle, try using fold(1) to fold the lines so you can see the assembler code. 5.6.3 Changes_to_the_C_Compiler o The types void and void * have been implemented. o The type qualifier const has been implemented. o Function prototypes are much more fully implemented and checked. For example, a cast of a function-pointer to a prototyped function-pointer works correctly. Completion of incomplete types in function prototypes is not done in messy circumstances (where, for example, multiple declarations of one function exist but complete types are found only by unifying all the declarations). o The new cc option, -prototypes turns on type checking of function prototypes, and -noprototypes turns it off. The checking is off by default so code that compiled with Release 4D1-3.2 should compile without complaint with this release. The option affects checking of function prototypes, not code generation, so programs compiled either way should work identically. In checking prototypes, the compiler allows various constructs that are not formally equivalent to get by with a warning to allow old code to compile with the /usr/include files supplied with this release. Differences such as declaring a function as returning a char * at one place and a void * at another receive a warning message if and only if -prototypes is on. Prototype differences such as declaring an argument as both unsigned short * and short * receive a warning only if -prototypes is on. The compiler attempts to be specific about the type discrepancies in function prototypes. Wherever possible it reports the types in English. Wherever possible it also reports the argument number. If you assign one function pointer to another, the function prototype on the right is the function prototype used unless the right function pointer has no prototype, in which case the function prototype, if any, of the pointer on the left is kept. o The ANSI C convention that adjacent string literals become a single string (ANSI section 3.1.4, December 1988 draft) is implemented in this compiler. o For future compatibility, do not put text after #else or #endif unless the text is enclosed in /* */ comment delimiters. While this release ignores arbitrary text on such preprocessor lines, ANSI C considers text after #else or #endif an error unless it is inside a comment. o The -acpp option to cc has been added. When -acpp is used, an ANSI C preprocessor is used instead of the normal C preprocessor. Source code for acpp (copyrighted by the Free Software Foundation) is in 4Dgifts. See the source for terms of the copyright. 5.6.3.1 Changes_to_Error_Messages There are many new error messages. Some of them might be confusing. These in particular have proven to be troublesome: o "prototype:pointer to struct is different from actual: pointer to struct" A prototype and an actual argument must be pointers to the same struct to be compatible. It is not enough that the structs pointed to be the same size. o "unknown size: void has no size" Many operations on pointers imply taking the size of the item pointed at. void * does not point at any object, so operations requiring the size are illegal. void *mfunc(); /* illegal: pointer arithmetic */ mfunc(x,y,z)+ 3; /* legal: pointer comparison to 0 */ !mfunc(x,y,z); o "constant evaluation found questionable result" The full message is: constant evaluation found questionable result as a result of copy propagation or other optimizations. Constant evaluation left to run time for this case A warning like this can come from the optimizer. If, as a result of constant propagation or other optimizations, a value is computed that produces an overflow, the optimizer does not do the operation at compile time but leaves it to run-time. For example, (m << n) is undefined in C if n is greater than m's width in bits. If you take the meaning as ``do the shift'' the result is zero! If the optimizer discovers that n is greater than 31 (for m, a constant int whose value is known to the optimizer) as a result of copy propagation, the shift is left to run-time and the message above is emitted. 5.6.4 Changes_to_dbx o C enumerated types are printed much more fully. To see an example of this, try whatis x on some enumerated type x. o whatis tries to report on files and variables when they have the same name. If x is a file named x.c and a global variable x, whatis x reports both usages. o As of this release, the kernel (IRIX) marks any file being dbx'd as busy. Consequently the old practice of doing a recompile and relink followed by the dbx rerun command does not work unless you first rm(1) the executable, because ld is unable to write to the busy text file. The lack of a busy mark in previous releases was quite dangerous, because you could relink and then do a dbx cont command on text that had changed. This resulted in serious problems during the debugging session. o A great deal of information was added to the dbx manual page and help file /usr/lib/dbx.help. The information in these files might be helpful to you. o The ccall command can now handle variables in interactive calls to functions in the program being debugged. Interactive calls can now use variables as arguments. o The $nextbreak debugger variable now controls how the next command is executed. By default, next is done at full speed. If done in the middle of recursions, next stops the next time the function is recursively called. If that is not what is desired, set $nextbreak to 0 or 1 before issuing next. See the help file for further details. 5.6.5 Changes_to_Other_Tools o lint, cxref, ctags, cb, and cflow understand the new ANSI C keywords- const, volatile, signed- as well as function prototypes. These now work properly on any code that compiles with cc. o lint now checks printf and scanf calls for accuracy. o Bodyless if and else clauses are diagnosed by lint. o lint mentions possible precedence confusions in many more circumstances than in previous releases. o lint has the cc -prototypes flag turned on. It is advisable to compile code with cc -prototypes before running lint. o ld now accepts the -U option. -U makes references to undefined symbol warnings instead of errors. This is useful in certain very large programs where leaving out the code for certain functions saves significant link time and where, for debugging purposes, those functions are not executed. Attempting to execute a missing function results in a segmentation violation. o Several changes have been made to make(1): - The new command line option -u has been added that causes all specified targets to be built unconditionally. - The sinclude directive has been added. sinclude works the same way that include does in a makefile, except that if the file specified in the sinclude command is not readable, the error is silently ignored by make(1). - make(1) now understands a convention similar to the alternate interpreter feature of exec(2). If the first line of a make command file starts with a string of the form "#!alternate_make", then make will attempt to invoke the specified alternate make with the same environment and arguments. - The function of the VPATH macro is now documented in the make(1) manual entry. - The behavior of VPATH has been enhanced to allow dependency rules to reference either the virtual target or the actual target. For example, either of the following may be specified: foo.o:foo.h $(VPATH)/foo.o:foo.h - The interaction between dependency rules and single suffix rules has been corrected. In previous versions of make(1), the presence of a dependency rule of the form cat:stdio.h caused the single suffix rule for .c: to be ignored. In the Release 4D1-3.3 version of make, single suffix rules are no longer ignored in this case. - The sizes of many of the internal tables used by make(1) have been increased or made dynamic to support long file names and very complex make command files. o Several changes have been made to pmake(1): - .f (Fortran) files were added to RCS transformation rules. - the new -u option (``unconditional'') remakes the specified targets (just like make(1)). - pmake now understands the sinclude directive (just like make(1)). - pmake now sets the MAKEFILE environment variable. It also puts command-line variables into the environment. - several makefiles can be specified on the command line with multiple -f options. - the new special targets .NOTPARALLEL, .ORDER, and .SINGLESHELL have been added and are described in the manual page. - pmake does not support rules for checking out SCCS files. It also ignores make(1)'s MAKEFLAGS environment variable. 5.7 Changes_to_the_Software_Installation_Tools The following are the major changes to the software installation tools for the 4D1-3.3 release: 5.7.1 Changes_to_Inst o The Manual Installation menu defaults item has been changed to standard. The functionality has not changed. o The recalculate item has been added to the Manual Installation and Subsystem Selection menus. It serves to initiate recomputation of the disk space requirements. A warning is given in this regard after shell escapes. o The Subsystem Selection menu has been reordered. o The yes item of the Selection menu has been changed to install, and the no item has been split into keep, for no action, and remove, to remove any old version that might be present. The remove requests are now queued along with install requests, rather than being executed immediately. o Removal requests may be made for installed products that are not part of the current distribution listing. These requests will be displayed by the list item. o The install item of the Selection menu has been changed to go, which serves to initiate the queued installation and removal requests. o The versions menu item serves to directly invoke the functionality of the versions command from within inst. The versions command may still be invoked as a separate IRIX command. o The sh and shroot items may no longer be invoked from the Interrupt/Error menu. o Many error messages have changed. o There is a dramatic increase in the speed of certain network operations. o The reliability and robustness of the on-line installation history updating mechanisms have been improved. There is no longer a separate recovery operation; retrying a failed installation or removal will usually correct the history. o There is now a check for incompatible subsystems being installed on the disk. The installation tool will not permit you to quit until this problem is resolved. 5.7.2 Changes_to_Versions o Versions is now built into inst, and can be invoked as a menu item at any time. o New products are being produced with dates where old ones had internal version numbers. The -n option can be used to display version numbers in the old form. o The -q option prevents all output, returning only exit status. o The format of the display has been improved. 5.8 Changes_to_4Dgifts As in previous system software releases, Release 4D1-3.3 includes a set of on-line source code examples on the DEV (development) tape in the dev.sw.giftssrc subsystem. These gifts are not installed or updated by default. To use them, you must first manually install them using inst(1M). For detailed information about using inst and the manual installation features, see Chapter 2, ``Installing Software''. Once the gifts are installed, you will find these source code examples in a sample user account directory called /usr/people/4Dgifts. 4Dgifts is set up as a sample user account to allow you to learn by example and to allow you to be productive while you are learning to use and customize the NeWS environment. You should find a README file in virtually every directory in and including ~4Dgifts. The README files describe the contents of each directory and provides information that will help you locate the types of source examples you are looking for. Read the file /usr/people/4Dgifts/README to for an explanation complete understanding of the contents of 4Dgifts. The source examples provided include programs demonstrating: o Fortran and C graphics o the fontmanager o how to use peripherals (e.g., the dial and button box) o the Hitachi digitizer (tablet) o a simple generic SCSI device driver o some cps, network, UNIX, and video examples o the January 1989 version of Kermit from Columbia University (This is provided only as a courtesy - Silicon Graphics does not support Kermit. See the README file in the kermit directory.) o a complete set of Silicon Graphics image libraries for creation and manipulation of image files o programs demonstrating a sample WorkSpace environment 5.8.1 New_4Dgifts_Directories New directories of note are: o ~4Dgifts/examples/audio An audio program for 4D/20 and 4D/25 Personal Iris machines demonstrating how to digitally record microphone input, place the data in a file, as well as playing back the sounds recorded. o ~4Dgifts/examples/glpg This subtree contains on-line versions of the source code examples given in the Graphics Library Programming Guide, Version 2.0 (with the exception of Chapter 18). o ~4Dgifts/examples/network This directory replaces the previous incarnation, tcp, and demonstrates how to use 4.3BSD Internet and UNIX domain sockets on IRIX. o ~4Dgifts/src/acpp Contains a set of files that comprise the ANSI C compatible pre-processor accessible via the cc(1) flag -acpp. o ~4Dgifts/src/dglfax Contains source code for the dglfax(1) utility. o ~4Dgifts/src/sphere Contains source code for the molecule demo /usr/demos/bin/mview, which makes use of libsphere.a, to show how functions in the sphere library are used. o ~4Dgifts/tutorials Contains interactive graphics tutorials previously only available to those who have attended SGI's 4D Graphics and Advanced Graphics courses. Areas covered include color, light, basic drawing primitives, and modeling and projection transformations. All tutorials are implemented with a user interface that allows you to interactively change the different parameters in real- time and see the effects these changes generate. o ~4Dgifts/wspace Contains WorkSpace goodies including: a subdirectory containing example .ftr files and scripts, a file containing ``poweruser'' short cut capabilities. 5.8.2 4Dgifts_Subdirectory_Descriptions The main subdirectories in 4Dgifts are: examples, kermit, iristools, src, and tutorials. The examples directory contains various subdirectories housing an assortment of code examples: ___________________________________________/No man is a failure who has friends Trevor Paquette ICBM:51'03"N/114'05"W|I accept the challange, body and soul, {ubc-cs,utai,alberta}!calgary!paquette|to seek the knowledge of the ones of old paquette@cpsc.ucalgary.ca | - engraved on the Kersa Blade of Esalon   Received: from vmb.brl.mil by VMB.BRL.MIL id aa01694; 25 Jun 90 22:47 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa01515; 25 Jun 90 22:26 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01478; 25 Jun 90 22:10 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa11408; 25 Jun 90 22:02 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA17634; Mon, 25 Jun 90 18:58:21 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 01:09:59 GMT From: Trevor Paquette Organization: The University of Calgary Subject: Irix 3.3 Changes Part 2/2 Message-Id: <1990Jun26.010959.24720@calgary.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Fortran graphics programs written in Fortran audio code demonstrating how to digitally record microphone input to a Personal Iris (only works on 4D/20 and 4D/25 machines), and play back these sounds cps four separate subdirectories each with a different type of example implementing the C to PostScript interface devices digitizer, dial and button box programs, as well as a program that uses the /dev/scsi generic SCSI driver fontmanager includes sample programs demonstrating usage of the Fontmanager Library, libfm.a glpg on-line versions of all source code examples contain in Chapter 18 of the Graphics Library Programming Guide, Version 2.0. grafix various C graphics programs nurbs contains four NURBS sample programs: one written in C, one for the DGL, one in Fortran, and one in Pascal network has some sample network programs using Internet (TCP/UDP) and UNIX domain sockets, as discussed in Chapter 3 of the Network Communications Guide. This directory was called tcp in previous releases. trackball contains four components of code for a Virtual Trackball Implementation: routines to calculate the virtual trackball, event-queue handling, drive a user-interface, and a simple program to use the other three unix contains fundamental examples of system programming video contains programs demonstrating usage of various video modes The kermit directory holds the public domain source and documentation for kermit, a file transfer protocol that is useful when you need to send files to and from an IRIS/UNIX computer and non-UNIX configurations such as VMS or DOS- based machines. The iristools directory contains a superset of special image libraries, image processing utilities, and graphics tools that used to exist, in the IRIS 3000 Series computers, under /usr/people/gifts/mextools. This source was used to build the binaries that now reside in /usr/sbin (e.g., cedit, showmap, ipaste, mag). In other words, every executable in /usr/sbin with a source file under iristools was built from that exact source (including the two libraries libgutil.a and libimage.a under iristools). The src directory contains an ANSI C preprocessor in "acpp", a "dglfax" subdir containing the source code for dglfax(1), and the directory "sphere" which houses the source for mview (the molecule demo) which shows how functions in the new sphere library--libsphere.a--are used. The tutorials subdirectory contains a set of interactive graphics tutorial programs that cover such areas of the GL as color, light, ortho2, projection, basic drawing primitives, and modeling transformations. All of these tutorials are implemented with a user interface that enables one to interactively change the different parameters in real-time and see the effects these changes generate. The directory /usr/people/4Dgifts is setup to work as a sample NeWS user login account replete with many template .ps files to help you understand the extent to which you can customize the NeWS environment. Along with a more substantial user.ps file, there is a startup.ps file, as well as a subdirectory /usr/people/4Dgifts/.4sight. This directory contains nine additional startup.ps files that show you how to create your own user-defined icons, window colors, menu fonts, et cetera. These files have comments throughout them to help describe what they do. There are many ways you can change and alter all of the possible startup variables. 5.8.2.1 Special_Gifts ./iristools/imgtools/snapshot.c This program allows you to interactively grab part or all of an image on the screen and dump it into an image file. It is the next generation of icut. By default, it is loaded in /usr/sbin as a gltool. See snapshot(6D). This image file can then be put back up on the screen with ipaste(1G), or sent to a supported printer with lp(1). ./{.workspace/*, README.wspace} There is an initial setup for a version of workspace with 4Dgifts that resides in the directory .workspace. The file README.wspace describes more of what is currently included. ./examples/grafix/{zrgb.c, zrgbmenu.c, zcmapmenu.c} ./examples/Fortran/{zrgb.f, zrgbmenu.f, zcmapmenu.f} These programs demonstrate aspects of z-buffering in various implementations. Of particular note are the zcmapmenu versions, which include a powerful example in the main infinite loop of how to write code that does not eat up extra CPU cycles (provided you do not need the animation to continue when the input focus is elsewhere) ./examples/devices/{iisc.c, inquire.c} Two programs that use the /dev/scsi generic SCSI driver. Be sure to also consult the README file in this directory. ./examples/audio/audio.c This program demonstrates how to digitally record microphone input on a Personal Iris (4D/20 or 4D/25) into a file, as well as playing back the sounds already recorded. ./examples/cps/{TagTestc, envdesign, item, vinews} These four directories contain examples that use the C to PostScript (cps) interface. 5.8.3 Installing_the_Gifts To install the 4Dgifts, log in as root, type inst, and follow the instructions on your screen. Refer to Chapter 2, ``Installing Software'', for a detailed discussion of inst. Choose the manual installation features, and explicitly select subsystem dev.sw.giftssrc. Once you have specified that you wish to use the manual installation features, type: select from the "Manual>" menu and then enter install dev.sw.giftssrc Now 4Dgifts will be included when you run the go menu item. The size of this account (uncompiled) is approximately 7133 blocks or about 3.56MB. 5.8.4 Setting_Up_4Dgifts_as_a_User_Login_Account Upon successful completion of loading the dev.sw.giftssrc subsystem from the Development tape (see below), you need to perform one more modification in order to set up /usr/people/4Dgifts as its own account: 1. Log in as root. 2. Edit the file /etc/passwd. Duplicate the ``guest'' passwd line. Change every occurrence of the word ``guest'' on this duplicate line to ``4Dgifts''. 3. Write the changes and exit the editor. 4. Now log out of the console screen entirely and log in as 4Dgifts. Startup is different from when you log in as guest, root or any of the other ``default'' login accounts. The intent here is that you copy ~4Dgifts/{.4sight, user.ps, startup.ps} into your home directory and try changing whatever parts you wish to make it place and define things more in the way you prefer. 5.9 Changes_to_WorkSpace The 4D1-3.3 release includes these changes to the IRIS WorkSpace: o Kernel change imon no longer blocks on select. Any changes made on the host machine appear immediately, rather than waiting one to three seconds. o fam When fam is invoked from inetd.conf, it can be passed an argument which determines the polling period when it is watching files over NFS. After the new argument is put in inetd.conf, you can activate it by becoming root and entering killall -HUP inetd. The programatic interface to fam was changed so that any fam request must be accompanied by the uid and gid of the calling process. fam then examines the file system with those settings, reverting to root privelege when it is finished. fam correctly reports when a directory disappears. o Sticky directories WorkSpace now deals with the problem of sticky directories. (These are directories which, if they are owned by another user, do not allow you to remove files owned by other users, and on which you don't have write permission, even though you have write permission in that directory). Note these behaviors: o If no sticky directories are involved, file transactions are generally as they used to be, although they should be much faster for large directories moved around on the same disk. o If a sticky directory owned by someone else exists and contains another user's file, and you do not have write permission for that file, 1. You cannot remove that file. 2. You cannot drag that file to a new directory. 3. You cannot remove its parent directory (even if you have full write permission on it). 4. If you have write permission on the parent, you can drag it to another directory on the same disk, but you cannot drag it across disks. These behaviors apply even if the files were dragged to the WorkSpace. o NFS If files are exported without root permissions, root explicitly cannot read them from another machine, unless there is read permission for others. Even if you own a directory on another machine, you cannot mount that directory and list its contents if you are running as root. fam now runs with the uid and gid of the calling process, so that you can run WorkSpace, visit remote mounted directories that you own, and view their contents. The 4D1-3.3 release includes this enhancement to IRIS WorkSpace performance: WorkSpace now updates itself only when all of the fam messages for a given burst are read off the select. This speeds up the opening of large windows. The 4D1-3.3 release includes these improvements to IRIS WorkSpace usability: o Name length of dumpster is no longer limited. You can rename the dumpster via the prefsheet, with a long name. o If directory view windows are stowed when WorkSpace is shut down, they now come up stowed when WorkSpace is restarted. Note that the blue WorkSpace window should, however, never come up stowed. o A stowed directoryView is no longer painted after update events until it is opened, so that however many update events occur, it paints only when it is opened.) o Arrange as group on WorkSpace should be a no-op if nothing is selected. o If you drag a file to a directory, and it overwrites the existing one, the new file now appears where you drag it to, instead of overwriting the existing one. o If dirview is requested to open a directory and is unable to do so, it now puts up a notifier naming the file and stating that it can't be opened. o When you try to do a file transaction and fail, the resultant notifier informs you of the name of the file it failed on. The 4D1-3.3 release includes these visual improvements to WorkSpace: o WorkSpace no longer leaves little white lines where the text view was if you move out of a scrolled view. o The text view no longer repaints itself briefly in the wrong place if you switch from view as list to view as icons or vice versa. o When an icon is dragged, the icon and the text field now paint themselves at the same time, instead of letting the text field paint first. o When you run as root for the first time (without a .workspace directory), the WorkSpace window is now layed out in a more attractive fashion. o When you tried to copy a file that was running, WorkSpace formerly created a file on the WorkSpace window, showing the icon as running. Now the new file is shown in its proper state. o WorkSpace now allows you to make a linked copy of the root folder. o When you are running with template and ask to restore root, it now gets its connection line immediately, instead of waiting until you move it. o getinfo windows now have no close on the menu. o If you run as root, the name of your dumpster is now /dumpster, not //dumpster. o When a window with cached data appeared, some icons formerly appeared before others. Now they all appear together. 5.10 Changes_to_4Sight The 4D1-3.3 release includes this change to 4Sight: o 4Sight was changed so that whenever a GL window is exited, the cursor color is reset to red. This was done so that programs that modify the cursor color won't interfere with other programs that work better with the default red cursor. However, the implication is that if a program does change the cursor color, it must do slightly more than what it did in the release 4D1-3.2. Now, to change the cursor color, a GL program must do the setting of the cursor color every time it receives an ``enter'' INPUTCHANGE event. ___________________________________________/No man is a failure who has friends Trevor Paquette ICBM:51'03"N/114'05"W|I accept the challange, body and soul, {ubc-cs,utai,alberta}!calgary!paquette|to seek the knowledge of the ones of old paquette@cpsc.ucalgary.ca | - engraved on the Kersa Blade of Esalon   Received: from vmb.brl.mil by VMB.BRL.MIL id aa01828; 25 Jun 90 23:08 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ad01515; 25 Jun 90 22:26 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab01483; 25 Jun 90 22:10 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa11424; 25 Jun 90 22:04 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA17552; Mon, 25 Jun 90 18:56:44 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 01:07:24 GMT From: Trevor Paquette Organization: The University of Calgary Subject: Irix 3.3 Major_Enhancements_and_Compatibility Message-Id: <1990Jun26.010724.24482@calgary.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL ------------------------- Irix 3.3 Release Notes 3. Major_Enhancements_and_Compatibility Taken from the command 'pcat /usr/relnotes/standard/ch3.z' and edited to remove underlining etc.. ------------------------- 3. Major_Enhancements_and_Compatibility This chapter describes the major software and hardware enhancements included in the 4D1-3.3 release, summarizes compatibility issues for the program development tools, and provides a description of the subsystems provided with this release. This chapter is divided into the following sections: o Major Enhancements o Compatibility of Program Development Tools o Description of 4D1-3.3 Subsystems 3.1 Major_Enhancements This section is divided into two subsections that describe the major software and hardware enhancements (respectively) in release 4D1-3.3. The major software enhancements in the 4D1-3.3 release include: o IRIX enhancements o New compiler features o 4Sight enhancements 3.1.1 Enhancements_to_IRIX Enhancements to IRIX fall into these categories: o POSIX o Virtual memory/file systems o Real-time programming o Parallel programming o Logical volumes In release 4D1-3.3, IRIX provides the ability to graft several physical disk partitions together into one logical partition, referred to as a logical volume. A logical volume can span multiple physical disks, providing both larger data files and greater file system throughput. Improved throughput is the result of increased parallelism in the I/O process by striping accesses across several drives and controllers. Using logical volumes, it is possible to create file systems as large as 8 gigabytes and individual data files as large as 2 gigabytes. Refer to the manual entries for mklv(1M), lvinit(1M) and lvtab(4) for more detailed information about logical volumes. A step-by-step procedure for creating a logical volume is given in the IRIS-4D System Administrator's Guide. o Extensible file systems The IRIX Extent File System (EFS) now allows partitions containing EFS file systems to be extended without bringing the system down or rebuilding the existing file system contained in the partition. For example, a logical volume containing an EFS file system can be extended by adding one or more disk partitions to the end of the existing logical volume. Note: File systems can only be extended if a logical volume is being created or if there is room remaining in the disk partition (this is usually not the case). After this has been done, the growfs(1M) utility program can be used to extend the existing EFS file system to fill the newly enlarged logical volume. This extension operation automatically preserves all the data in the original file system. It is not necessary to mkfs(1M) and then reload the data. o POSIX compliance As a beta site for the NIST POSIX Verification Suite, Silicon Graphics is committed to providing a POSIX- compliant operating system. The 4D1-3.3 release provides a major step towards POSIX (IEEE 1003.1) compliance. All functions and interfaces specified by the 1003.1 Full Use Standard are provided with the following exceptions: o The behavior of file system operations specified by 1003.1 is guaranteed only for local EFS file systems, not for NFS file systems. o The ANSI setlocale function as specified and extended by Chapter 8 of the 1003.1 spec is not implemented. o The POSIX tar and cpio program extensions specified in Chapter 10 are not provided. Note: At the time release 3.3 was completed, the NIST suite had not yet been released in final form, so Silicon Graphics cannot warrant that release 3.3 will pass the final released version of the NIST POSIX Verification Suite. o Merged page and buffer caches A number of significant changes have been made to the kernel algorithms that manage system memory. All system memory is now treated as a large cache that holds pages of the virtual space of processes and pages from disk files. These changes have a number of beneficial effects, including significant performance improvements in the file system and correct semantics when a given file is accessed both as a mapped file and through the traditional read/write interface. The cache is subordinate to memory attached to user processes, so caching activity will not affect currently running processes. Any page of free memory, however, is now available to cache disk data, as long as the file's corresponding inode is resident in the in-core inode cache. o Resident set size limitation The kernel page replacement algorithm has been changed to provide the ability to limit the number of physical memory pages that a particular process can use at any given time. This is called the Resident Set Size (RSS) of a process. By using this feature, it is possible to minimize the effect that background jobs doing heavy paging have on the interactive response of the system. Refer to the setrlimit(2) manual entry for further information. o No more preallocation of swap space In previous versions of IRIX, it was impossible to create any virtual space without having a place to store it (i.e., main memory or swap space). In Release 4D1-3.3, the system has been changed so that storage space is required only when the virtual space actually needs to be stored. For example, when a large program calls fork(2) and then immediately calls exec(2) to execute a smaller program, it is no longer required that the system have enough swap and memory to hold twice the size of the parent process. Since fork(2) marks the pages of the parent copy-on-write and most of the parent's pages are never touched by the child before it calls exec(2), separate storage space is not required for most of the child's logical pages in the interim between the fork and the exec. o Autogrowable mapped files In this release it is possible to extend a mapped file by storing to a virtual address corresponding to a location beyond the current end of file. At the time the file is originally mapped using the mmap(2) system call, the caller specifies the file offset and length of the segment to be mapped. If the specified segment reaches beyond the end of file and autogrow mode has been requested for the segment, then stores to virtual addresses that correspond to locations beyond the end of file cause the file to be extended automatically to the required length. The extension is done in multiples of pages (4096 bytes each) up to the limit specified by the offset and length given at mmap(2) time. All pages added to the file will be zero filled. Refer to the mmap(2) manual entry for more information. o Large virtual spaces handled more efficiently The performance of programs that have large virtual address spaces (more than 10 megabytes of code or data) or many virtual address spaces (using many shared libraries or mapped files) is much better in Release 4D1-3.3, as a result of changes in the way the kernel virtual memory code manages the Translation Lookaside Buffer (TLB). The TLB is a 64-entry cache on the microprocessor that caches virtual to physical address translations. Large programs that exhibit poor reference locality generate significantly fewer TLB faults in Release 4D1-3.3, resulting in much better performance. 3.1.1.1 Real-time_Programming The IRIX operating system supports a powerful set of real-time programming features. In combination, a programmer can use these features to time events accurately, use signal handlers as true interrupt routines, control allocation of real memory to the process and provide for priority scheduling. Real-time programming provides for a range of response time and latency from very fast handling at device interrupt time to high-priority dispatch of user processes to handle the event. The following enhancements have been made to the real-time features of IRIX in Release 4D1-3.3: o High-resolution interval timers and time-of-day clock, accurate to 833 microseconds. The maximum resolution of the system time-of-day (as returned by the gettimeofday(2) call) and interval timers (see setitimer(2)) has been increased from 10 milliseconds to 833 microseconds. By default, the resolution is set to 10 milliseconds. If you want the greater resolution, you can enable it as the superuser with the new ftimer(1) command. o Lower interrupt latency The 4D1-3.3 release guarantees a maximum interrupt latency of 650 microseconds on properly configured systems. Maximum interrupt latency has been decreased by reducing the worst-case time that interrupts are masked by the operating system. On properly configured systems, worst-case interrupt latency is 650 microseconds (assuming a 25 MHz processor). o Kernel preemption points have been added to allow non- degrading priority processes to preempt lower priority processes even when the preempted process is executing in the kernel. This provides a significant improvement in worst-case process dispatch latency. 3.1.1.2 Parallel_Programming o Gang scheduling Cooperating processes can now be scheduled as a unit. This ``gang scheduling'' of processes on multiple- processor systems greatly increases the efficiency with which cooperating processes communicate with one another. With gang scheduling it is possible to run parallel jobs efficiently in a time sharing environment. This results in better overall system performance, particularly on systems supporting many simultaneous users or tasks. See schedctl(2) for more information. o Shared graphics within a parallel job In Release 4D1-3.3, it is possible for graphics to be a shared attribute within a shared process group. In previous releases, only one process within the share group could be a graphics process. A process that is already a graphics process (that is, has called winopen(3G)) and calls sproc(2) with a share mask that specifies the sharing of address space automatically creates a shared process group in which graphics is shared. This means that any process within the share group can make Graphics Library calls and that all such calls affect the same window or graphics context. It is important to note that Graphics Library code does not prevent separate processes within a parallel job from simultaneously accessing Graphics Library data structures or the graphics pipe. Because the programming model presented by the Graphics Library is fundamentally stateful, the responsibility for the definition and protection of critical regions must be owned by the application program. For example, suppose that two processes within the share group want to perform a bgnpolygon, v3f, v3f, ... endpolygon sequence, each for a different polygon. As soon as one of the processes has done its bgnpolygon, the other process must not touch the pipe until the first process has done the corresponding endpolygon. Thus the application code must make the above sequence a critical region in each process in order to ensure that the two processes do not interleave their sequences of calls. o Improvements to signal handling in parallel jobs In Release 4D1-3.3 it is possible to generate a signal to all members of a shared process group when any member of that group dies. The prctl(PR_SETEXITSIG) system call is used to activate this behavior and to specify the signal that is delivered. Refer to the prctl(2) and sproc(2) manual entries for details. o Blocking and unblocking entire shared process groups Two new routines are provided that are analogous to blockproc(2) and unblockproc(2) but that act on entire shared process groups. These are blockprocall(2) and unblockprocall(2). Refer to the blockprocall(2) manual entry for details. o New debugging support in libmpc Two new routines have been added to aid in debugging parallel programs that use the locking primitives in libmpc: usdumplock(3) and usdumpsema(3). These primitives provide the ability to dump the history associated with a single lock or semaphore, respectively. 3.1.2 New_Compiler_Features o The 4D1-3.3 release integrates MIPS 2.0 compiler technology. o The new compilers contain numerous improvements and fixes over the previous release, most notably in the code optimizer. o Nearly complete ANSI C support: o Include files provided with this release have been changed to use ANSI C-type prototypes. o ANSI rules and type checking can be turned on with a compiler switch. o New ANSI types (for example, void, void*, const) and ANSI constructs are recognized. For developers who do not need to conform to the ANSI standard immediately, the new compiler permits non-ANSI constructs recognized by the current compiler. If you are porting C applications from Release 4D1-3.2 to 4D1-3.3, be aware that the ANSI C compiler outputs numerous new warning messages where the previous compiler remained silent. If ANSI compatibility is not required for applications you are developing under Release 4D1-3.3, you can ignore these error messages. For more information on the specific ANSI support introduced in Release 4D1-3.3, read ``Changes to Program Development Tools'' in Chapter 5, ``Changes''. 3.1.2.1 4Sight_Enhancements Release 4D1-3.3 includes a new version of 4Sight (1.5) in which retained canvases are implemented when 4Sight is running in 24-bit RGB mode. Previously retained canvases only worked when the server was running in 8-bit color-index mode. A backgammon program has been included in /usr/NeWS/demos as a demonstration of retained canvases. Psview(1) no longer caches the current page in 4Sight. This means that the PostScript programs that use the currentfile can now be previewed. It also means that 4Sight no longer grows during extensive previewing. A complete and accurate manual page is now provided. Many bugs have been fixed as described later in these release notes. 3.1.3 Graphics 3.1.3.1 Sphere_Library Release 4D1-3.3 includes a new subroutine library called the Sphere Library, that draws spheres by issuing GL calls. It can be called from C or FORTRAN programs. The Sphere Library files are /usr/lib/libsphere.a and /usr/include/gl/sphere.h. There is an example program located in /usr/people/4Dgifts/src/sphere. Appendix B of these Release Notes contains manual pages for the Sphere Library routines. These manual pages are also available on-line. 3.1.4 Hardware_Enhancements The 4D1-3.3 release supports the following new Silicon Graphics hardware: o PowerVision Graphics o 4D/300 POWER Series systems o POWER Channel IO controller for POWER Series systems o IPI2 disk subsystem for POWER Series systems o System memory expansion to 256 MB on POWER Series systems o RV2 Raster Subsystem o 14" Monitor Support (for 4D/20 and 4D/25 systems) o Entry Plus Graphics Upgrade (for 4D/20 and 4D/25 systems) o Multihead Personal IRIS 3.1.4.1 PowerVision_Graphics The PowerVision system is a graphics supercomputer that renders 1 million polygons per second and includes hardware support for real-time texture mapping, polygon antialiasing, fog/haze effects, sub- millisecond full-screen clear, arbitrary clipping planes, and image processing operations. It also includes hardware support for near-real-time rendering of higher-quality images, with full-scene antialiasing, depth of field, and motion blur. Programs compiled with gl_s will run on the VGX workstations without recompilation. PowerVision is available as a complete system and as an upgrade to existing members of the IRIS Power Series product line. 3.1.4.2 4D/300_POWER_Series_Systems These multiprocessor systems feature CPU boards which have 33 MHz microprocessors, floating point units and associated caches. Each 33 MHz processor board contains two complete CPU and FPU pairs. This board is also available as an upgrade to existing 4D/200 series systems in both server and graphics configurations. 3.1.4.3 POWER_Channel_IO_Controller The POWER Channel IO Controller provides a significant increase in IO bandwidth for POWER Series systems. The POWER Channel has the following features: o Up to 32 megabytes per second VME throughput o Two complete SCSI controllers, both supporting synchronous SCSI o Ethernet controller with DMA capability 3.1.4.4 IPI2_Disk_Subsystem IPI2 is a new high performance VME disk interface. Each IPI2 controller can support up to 8 drives on each of its two independent channels. Each drive has an unformatted capacity of 1.1 gigabytes and a raw transfer rate of 6 megabytes per second. The largest configuration supported at first release will be two controllers, supporting 16 drives each, for a total of 32 gigabytes of high performance disk capacity. 3.1.4.5 256_MB_Memory Through the use of 4 megabit dynamic RAMs, POWER Series machines now support a maximum memory configuration of 256 megabytes. 3.1.4.6 RV2_Raster_Subsystem The 4D1-3.3 release includes support for the RV2 raster subsystem. The RV2 is a single board replacement for the RV1/RM1 three-board set. The GT/GTX systems shipped after 7/1/90 may include RV2 boards and thus require the 4D1-3.3 release. Note: The RV2 raster subsystem does not contain alpha bitplanes. 3.1.4.7 14-Inch_Monitor_Support The 14-inch monitor differs from our standard 19-inch monitor in that its resolution is only 1024x768 instead of 1280x1024. This difference is automatically recognized by the system software, and the default transforms for graphics are set accordingly. However, there are times when applications need to know the size of the monitor, for example, when placing and sizing windows. There are getgdesc inquires to return this information (see getgdesc(3G)). _____________________________________________ getgdesc Inquiry Return Value ______________ 19" Monitor 14" Monitor ____________________________________________ GD_XPMAX 1280 1024 GD_YPMAX 1024 768 GD_XMMAX 340 240 GD_YMMAX 272 180 _____________________________________________ 3.1.4.8 Entry_Plus_Graphics_Upgrade The Entry Plus graphics upgrade allows Z buffered RGB images to be displayed on an eight bit Personal IRIS. This upgrade includes a Z buffer and overlay planes. RGB images are displayed in 8 bits by using dithering to approximate the true color. When RGB mode is enabled in this configuration, double buffering is turned off (because 4 bit RGB mode is not supported). The getgdesc call can be used to get information about the configuration of the current system. In the case of an Entry Plus graphics configuration, getgdesc will return the following values: _____________________________________ getgdesc Inquiry Return Value ____________________________________ GD_BITS_NORM_SNG_RED 3 GD_BITS_NORM_SNG_GREEN 3 GD_BITS_NORM_SNG_BLUE 2 GD_BITS_NORM_DBL_RED 0 GD_BITS_NORM_DBL_GREEN 0 GD_BITS_NORM_DBL_BLUE 0 _____________________________________ GD_BITS_NORM_SNG_CMODE 8 GD_BITS_NORM_DBL_CMODE 4 _____________________________________ This information indicates that there are three bits of red, three of green, and two of blue in single buffered RGB mode. It also shows that double buffered RGB mode is not supported. 3.1.4.9 Multihead_Personal_IRIS For applications that demand multiple screens, the Multihead Personal IRIS is now available. This workstation incorporates up to four multiple Personal IRIS graphics cards, each with their own monitor. The system is based on the IRISette VME card set and is packaged in a Professional IRIS chassis. In this release, only two screens are supported, the second of which is without window management. 3.2 Compatibility_of_Program_Development_Tools This section describes the compatibility between these software releases: 4D1-3.2 and 4D1-3.3 Release 4D1-3.3 is generally compatible with release 4D1- 3.2. 3.2.1 Upwards_Compatibility Binaries linked with the non- shared versions of libfm.a or libgl.a must be relinked. Silicon Graphics recommends always linking with the shared versions of the font manager and the GL. Most system header files have been changed to conform to the ANSI C specification; in particular, non-ANSI names have been renamed and function prototypes added. As a result, some previously undetected "incorrect" C programs will be detected. There are a large number of enhancements and changes to Fortran. Please read the Fortran release notes for Fortran specific compatibility information. In the following table, source means source files, whether Fortran, C, or another language. Application libraries means not only archive libraries of object files (.o files), but also individual object files. Executables means executable binaries (not shell command files). _______________________________________________________________________ | | 4D1-3.3 System Software | | |______________________________________________________| |4D1-3.2 | Compilation | Tools | Execution | |Generated | Environment | Environment | Environment | | | (cc, f77, ar,| (nm, strip, | | | | ld) | dbx, edge, | | | | | lint, stdump, | | | | | size, dis) | | |______________|_______________|__________________|___________________| |source | yes | NA | NA | |______________|_______________|__________________|___________________| |application | | | | |libraries | yes | yes | NA | |______________|_______________|__________________|___________________| |executables | NA | yes | yes, unless | | | | | the non-shared | | | | | versions of | | | | | the font | | | | | manager or GL | | | | | were used | |______________|_______________|__________________|___________________| 3.2.2 Backwards_Compatibility While object files and executables created under 4D1-3.3 may work on an older system, this use is not recommended. Indeed, it is dangerous to run a new executable on an older system as it may have unpredictable consequences. The 3.2 loader, ld, will generate warning messages if it encounters an object file generated on a 3.3 system. This warning can be ignored. Due to changes in the shared library format, executables linked under 3.3 to shared libraries will not execute correctly under 3.2 or older systems. And, of course, executables linked to private libraries under 3.3 should not be executed on 3.2 or older systems. 3.3 Description_of_4D1-3.3_Subsystems 3.3.0.1 Execution_Only_Environment_Tape_1 eoe1.sw.unix This subsystem contains the core IRIX commands and files and is required on all systems. eoe1.man.relnotes This is an on-line readable copy of the IRIX release notes. eoe1.man.unix This is an on-line copy of the IRIX user's manual. Entries may be viewed with the ``man'' command. 3.3.0.2 Execution_Only_Environment_Tape_2 eoe2.sw.NeWS This is the 4Sight window manager and is required by all workstations. This subsystem contains the ``standard'' NeWS fonts. Courier Icon Courier-Bold Iris Courier-BoldOblique Screen Courier-Oblique Screen-Bold Cursor Times-Bold Helvetica Times-BoldItalic Helvetica-Bold Times-Italic Helvetica-BoldOblique Times-Roman Helvetica-Oblique type eoe2.sw.X11 This is the run-time X11 windowing package. It is required to run X11 applications. This subsystem includes the following fonts that are required to run X, along with some terminal emulator fonts expected by xterm users: 6x10 cursor 6x12 Terminal 6x13 Terminal-Bold 8x13 Terminal-BoldNormal 8x13b TerminalNormal 9x15 eoe2.sw.Xapps These are X client applications. eoe2.sw.Xdemos These are demonstration programs that make use of the X windowing system. You must install eoe2.sw.X11 in order for these programs to work. eoe2.sw.Xfonts These are a set of fonts which were shipped with X11 Release 2. They are no longer a part of X, but some older X programs require these fonts to be in place. apl German sup arrow Greek supsup chess Hebrew swd chp ipa sym cyr krivo vbee Cyrillic lat vctl dancer met vg fcor micro vgb fg mit vgbc fg-Bold oldera vgh fg-Oblique plunk vgi fgone rot vgvb fgone-Bold runlen vgl fgone-Oblique sansserif vmic fgs sansserif-Bold vply fqxb sansserif-Oblique vr fr serif vrb fr-Bold serifb vri fr-Oblique serifi vsg frone stan vsgn frone-Oblique stempl vshd frthree sub vxms frtwo subsub xif eoe2.sw.acct This is the System V Process Accounting package. It is used to monitor system resource usage on a per-user basis. When Process Accounting is installed and turned on, a record is kept of every command that is executed along with data on the resources the command used. Report generating scripts are then run which produce periodic reports of system utilization. This package is useful when it is desired to monitor usage patterns in systems with large numbers of users. eoe2.sw.bsdlpr This subsystem contains 4.3BSD line printer software to allow the IRIS to access remote printers using the lpd protocol. Using the lpd daemon to control local printers is not supported. eoe2.sw.cdsio This is the driver for the VME Serial Expansion board. It is required for systems with that option installed. eoe2.sw.crypt This subsystem contains the crypt and makekey programs which, along with the editors ed and vi, implement a simple text encryption system. eoe2.sw.demos This is the standard SGI demonstration package. It has been enhanced and expanded with this release and features a menu driven front end called ``buttonfly''. Some programs may not image properly on machines with eight bitplanes or without a Z buffer. eoe2.sw.dfm This is a subset of IRIX utilities dealing with directory and file management. It is generally required in all systems. eoe2.sw.editors This subsystem contains the standard IRIX text editors and is generally required on all systems. eoe2.sw.envm This is the IRIX WorkSpace package, a user-friendly alternative to the standard IRIX command shell. eoe2.sw.gltools This subsystem contains a collection of simple tools that perform a variety of graphics-related functions on a workstation. Complete source code for these tools is contained in the dev.sw.giftssrc subsystem described in the next section. These tools include: blanktime imged mag cedit icut mousewarp clock imgexp savemap dialwarp interp scrsave gamcal ipaste showmap gamma istat snapshot ical loadmap textcolors eoe2.sw.hyper This is the driver for the HyperNet VME interface and is required only on systems with that option installed. eoe2.sw.ikc This is the driver for the VME Parallel Expansion interface and is required only by systems with that option installed. eoe2.sw.ipc This subsystem contains utilities useful in controlling the System V Inter Process Communications facility. It consists of two utilities: ipcrm and ipcs. eoe2.sw.ipgate This subsystem contains additional network routing daemons for machines with multiple network interfaces (gateways). The gated routing daemon supports RIP, EGP and HELLO routing protocols. The mrouted allows forwarding of IP multicast packets between networks. eoe2.sw.lp This subsystem contains the Line Printer Spooling package and is required on any system that will be used with printers, either locally or over a network. eoe2.sw.mast This subsystem contains vital graphics software and is generally required on workstations. On some systems, this subsystem will be empty however. eoe2.sw.moregltools This is an extension of the eoe2.sw.gltools subsystem and contains commands that implement simple image processing functions. The complete source to these commands is contained in the def.sw.giftssrc subsystem described in the next section. eoe2.sw.optfonts These are additional fonts that are not required by the base system but may be desirable for some applications. Boston NewCenturySchoolbook-Bold Charter-Black NewCenturySchoolbook-BoldItalic Charter-BlackItalic NewCenturySchoolbook-Italic Charter-Italic NewCenturySchoolbook-Roman Charter-Roman Symbol Kanji eoe2.sw.perf This subsystem contains the System Activity Reporting (SAR) package. This facility is useful for monitoring processor activity and IRIX system performance. It may be used in either an interactive mode or as a background data collector/report generator. This subsystem is not required, but may be useful in diagnosing system performance problems. eoe2.sw.spaceball This subsystem provides IRISphere support. eoe2.sw.spell This subsystem contains a dictionary and command that checks spelling. eoe2.sw.sysadm This is a system administration package that does not require graphics and is intended for use primarily on server machines. eoe2.sw.tcp This subsystem contains programs and files that implement the TCP/IP family of networking facilities. It is generally required on all systems even if they will be used in a stand-alone environment. eoe2.sw.terminf This is the Terminfo terminal database. It contains files describing the capabilities of hundreds of different types of terminals and is used by the vi editor as well as many common terminal- oriented applications. A few common terminals may be used without this subsystem. eoe2.sw.ts This is the driver for the ISI VME quarter-inch cartridge tape controller. It is required on systems with that option installed. eoe2.sw.uds This subsystem contains support for 4.3BSD UNIX domain sockets if you install eoe2.sw.tcp. This subsystem is required if you install the eoe2.sw.bsdlpr subsystem. eoe2.sw.usrenv This subsystem contains a subset of the standard IRIX commands and is generally required on all systems. eoe2.sw.uucp This is the traditional UUCP communications package which implements a point-to-point networking facility. This is only required for sites where UUCP is used but it also contains facilities that may be useful for systems with modems. eoe2.sw.vadmin This is a graphical interface to the standard IRIX administration utilities. It provides user-friendly tools for managing printers, users, disks, networks and other common administrative function. eoe2.sw.xm This is the driver for the Xylogics VME 1/2" tape controller and is required only on systems with the controller installed. eoe2.sw.xylsio This provides 16-port serial board support. eoe2.man.X11 This is an on-line version of the IRIX User's Reference entries that describe the commands in the eoe2.sw.X11 subsystem. eoe2.man.bsdlpr These are man pages for the BSD lpd/lpr line printer utilities. eoe2.man.demos This is an on-line version of the IRIX User's Reference, section 6, which describes the demonstration programs. eoe2.man.spaceball These are man pages that support the IRISphere product. 3.3.0.3 Development_Option dev.sw.G0libraries This subsystem contains versions of all of the standard programming libraries compiled with the ----GGGG 0000 option. These versions of the libraries are generally not required. dev.sw.NeWSimg This is an image database for the 4Sight image demonstration programs. dev.sw.bsdhdrs This subsystem contains symbolic links for BSD header files formerly in /usr/include/bsd. In Release 3.3, BSD header files are in /usr/include. This subsystem is needed if you are upgrading from previous releases or you have source code that include BSD headers with explicit mention of the bsd subdirectory, for example: #include dev.sw.cc This subsystem is contains all of the standard IRIX commands and files for compiling and debugging C programs. dev.sw.cedgetut This contains files that accompany documentation on how to use Edge, the IRIX window-based debugger, when programming in C. dev.sw.crypt This subsystem contains the file libcrypt.a for use by programs that perform data encryption. dev.sw.debug This contains the IRIX kernel debugger and is useful only to those developing kernel device drivers. dev.sw.giftssrc This subsystem contains a multitude of sample programs in source code form including the source for all of the eoe2.sw.gltools and eoe2.sw.moregltools subsystems described in the previous section. While the installation of this subsystem is not required, many developers have found the example programs to be extremely useful in learning about the GL, networking, and generic SCSI interfacing. dev.sw.moredemos This is an extension of the eoe2.sw.demos subsystem and contains more demonstration programs. These programs are segregated here due to disk space requirements. dev.sw.rcs This contains the Revision Control System (RCS) which is a set of programs that may be used to control a source code development project. With RCS, changes to source files are kept in a database with comments such that previous versions of a particular file may be retrieved. dev.sw.sccs This contains the Source Code Control System which is identical in purpose although different in use to RCS (described above). dev.man.cc This contains the IRIX Programmer's Reference manual in an on-line readable format. 3.4 Subsystem_Sizes This is a list of all the subsystems and their sizes. Default Install indicates subsystems that are are installed by default when you install using automatic mode, or you select default. To install subsystems that are not installed by default, you must select them with manual installation features. All sizes are in kilobytes. ___________________________________________________________________ subsystem Personal IRIS 4D Data Power Default IRIS & Station Vision Install Power Series __________________________________________________________________ eoe1.sw.unix 16248 15085 12081 11796 yes eoe1.man.relnotes 185 185 185 185 yes eoe1.man.unix 2234 2234 2234 2234 yes __________________________________________________________________ TABLE 1. S4-EOE1-3.3 _____________________________________________________________________ subsystem Personal IRIS 4D Data Power Default IRIS & Station Vision Install Power Series ____________________________________________________________________ eoe2.sw.NeWS 6565 6565 5949 6565 yes eoe2.sw.X11 2192 2192 2192 2192 no eoe2.sw.xapps 11602 11602 11602 11602 no eoe2.sw.Xdemos 1444 1444 1444 1444 no eoe2.sw.Xfonts 2368 2368 2368 2368 no eoe2.sw.acct 468 468 468 468 no eoe2.sw.bsdlpr 550 550 550 550 no eoe2.sw.cdsio 23 23 0 0 yes eoe2.sw.crypt 29 29 29 29 yes eoe2.sw.demos 2385 2274 1385 2277 yes eoe2.sw.dfm 499 499 499 499 yes eoe2.sw.editors 385 385 385 385 yes eoe2.sw.envm 3435 3435 3435 3435 yes eoe2.sw.gltools 425 425 0 425 yes eoe2.sw.hyper 99 99 99 99 no eoe2.sw.ikc 15 15 0 0 yes eoe2.sw.ipc 121 121 121 121 yes eoe2.sw.ipgate 334 334 334 334 no eoe2.sw.lp 759 759 759 759 yes eoe2.sw.mast 0 490 0 1528 yes eoe2.sw.moregltools 1307 1307 1307 1307 no eoe2.sw.optfonts 2273 2273 2273 2273 no eoe2.sw.perf 802 802 782 782 yes eoe2.sw.spaceball 399 399 399 399 no eoe2.sw.spell 433 433 433 433 no eoe2.sw.sysadm 260 260 260 260 yes eoe2.sw.tcp 4236 4236 4236 4236 yes eoe2.sw.terminf 956 956 956 956 no eoe2.sw.ts 15 15 15 15 no eoe2.sw.uds 18 18 18 18 yes eoe2.sw.usrenv 526 526 526 526 yes eoe2.sw.uucp 1126 1126 1126 1126 no eoe2.sw.vadmin 4425 4426 4426 4426 yes eoe2.sw.xm 31 31 31 31 no eoe2.sw.xylsio 28 28 0 0 no eoe2.man.X11 563 563 563 563 no eoe2.man.bsdlpr 23 23 23 23 no eoe2.man.demos 164 164 164 164 yes eoe2.man.spaceball 10 10 10 10 no ____________________________________________________________________ TABLE 2. S4-EOE2-3.3 ____________________________________________________________________ subsystem Personal IRIS 4D Data Power Default IRIS & Station Vision Install Power Series ___________________________________________________________________ dev.sw.G0libraries 5038 5038 3731 3731 no dev.sw.NEWSimg 6360 6360 6360 6360 no dev.sw.bsdhdrs 0 0 0 0 no dev.sw.cc 10607 10619 9292 8995 yes dev.sw.cedgetut 8 8 8 8 no dev.sw.crypt 15 15 15 15 yes dev.sw.debug 433 329 157 157 no dev.sw.giftssrc 3548 3548 3548 3548 no dev.sw.moredemos 3165 5675 2967 9634 no dev.sw.rcs 839 839 839 839 no dev.sw.sccs 845 845 845 845 no dev.man.cc 2062 2062 2062 2062 yes ___________________________________________________________________ TABLE 3. S5-DV01-3.3 ___________________________________________/No man is a failure who has friends Trevor Paquette ICBM:51'03"N/114'05"W|I accept the challange, body and soul, {ubc-cs,utai,alberta}!calgary!paquette|to seek the knowledge of the ones of old paquette@cpsc.ucalgary.ca | - engraved on the Kersa Blade of Esalon   Received: from vmb.brl.mil by VMB.BRL.MIL id aa01973; 25 Jun 90 23:34 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ab01515; 25 Jun 90 22:26 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01481; 25 Jun 90 22:10 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa11418; 25 Jun 90 22:03 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA17587; Mon, 25 Jun 90 18:57:18 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 01:08:21 GMT From: Trevor Paquette Organization: The University of Calgary Subject: Irix 3.3 Additions Message-Id: <1990Jun26.010821.24567@calgary.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL ------------------------- Irix 3.3 Release Notes 4. Additions Taken from the command 'pcat /usr/relnotes/standard/ch4.z' and edited to remove underlining etc.. ------------------------- 4. Additions This chapter describes additions to IRIX, the Graphics Library, and demos in the 4D1-3.3 software release. 4.1 Additions_to_IRIX 4.1.1 BSD_Compatibility Developers who are porting BSD-based applications to 4D1-3.3 will find the transition to IRIX easier with this release. Release 4D1-3.3 provides full 4.3BSD signal support as well as BSD header file and library compatibility. New BSD features supported in Release 4D1-3.3 include: o per-process multiple group IDs - The default under IRIX is that a process has one group ID but BSD multiple group mode can be activated with the multgrps(1) command. o the sigvec(3B) routine and all the associated calls for specifying signal handling o the setrlimit(2) system call for limiting process resource consumption o UNIX domain sockets o the BSD line printer spooler (lpr and lpd) IRIX now includes a version of the Berkeley lpr/lpd print spooler system. All of the Berkeley spooler commands are available, together with their manual entries. This software allows IRIS-4D systems that are configured in a heterogeneous network to send print requests to BSD print servers. Note: The functionality of this printer spooler is limited to the functionality provided in the 4.3BSD system. In particular, there is no support for PostScriptr printers. For this reason, it is not supported for use as the primary native spooler system for IRIS-4D systems. o the 4.3BSD versions of diff(1), chown(1), and chgrp(1). These commands are a superset of the System V versions. o the renice(1), w(1), uptime(1), and whoami(1) programs. o the 4.3BSD libc routines setlinebuf(3), setbuffer(3), vsyslog(3), getrusage(3), and vfork(2) have been added. vfork is emulated using sproc(2) and prctl(2). 4.1.1.1 BSD_Header_File_Relocation In this release, BSD header files formerly in the /usr/include/bsd subdirectory are now in /usr/include. Therefore, you no longer need the cc(1) directive -I/usr/include/bsd. For compatibility with previous IRIX releases, the dev.sw.bsdhdrs subsystem contains symbolic links for files that used to reside in /usr/include/bsd, linking them to their new locations. If you had (non-portable) source code that included BSD headers with the explicit bsd/ prefixes in the #include statement, it is worthwhile to remove the prefix for portability. Future releases of IRIX may not provide these links. 4.1.1.2 BSD_Library_Compatibility Most of the routines in /usr/lib/libbsd.a are now in the standard C library. These include routines such as gethostbyname(3N), rcmd(3N), and ndbm(3). Therefore, it is not necessary (in most cases) to use the cc(1) directive -lbsd when compiling and linking BSD programs. The remaining routines in libbsd.a have the same names as standard C library routines, but have different invocation syntax or semantics. These routines include: getgroups (3B) initgroups (3B) setgroups (3B) getpgrp (3B) setpgrp (3B) chown (3B) fchown (3B) dup2 (3B) The BSD versions of these routines are also in the standard C library with the prefix BSD (for example, BSDchown) in case you need to ``mix and match'' System V and BSD routines. Note that in previous releases there was no setpgrp routine in libbsd.a. Thus, a program that was linked with -lbsd and called setpgrp was actually calling the System V version of setpgrp in previous releases, but will call the BSD version when compiled the same way under Release 4D1-3.3. Since the syntax of the two versions of setpgrp is different, this change may cause some porting problems. Refer to the intro(3) manual entry for more details about the procedure for compiling and linking BSD compatible programs. 4.1.2 File_System_Additions As part of supporting the BSD/POSIX multiple group feature, Release 4D1-3.3 IRIX also provides the ability to give BSD group ownership behavior when creating files. When a process creates a file in a System V.3 system, the owner of the new file is the effective user id of the process creating the file and the group of the file is set to the effective group ID of the creating process. In a BSD system, the owner of a new file is set to the effective user ID of the process creating the file, but the group of the file is set to the group of the containing directory. The default behavior in Release 4D1-3.3 IRIX is the System V.3 behavior. You can specify the BSD behavior in two different ways: on an individual directory or on every directory in an entire file system. To specify the BSD behavior for all files created within a single directory, turn on the set group ID permission bit for that directory. You can do this with the following command: chmod g+s directory_name To activate the BSD behavior on an entire file system, set the grpid mount option on that file system at mount time by specifying the grpid mount option in the /etc/fstab entry for that file system. The following fstab entry does this: /dev/usr /usr efs rw,raw=/dev/rusr,grpid 0 0 Refer to the chmod(1) and fstab(4) manual entries for further information. 4.1.2.1 Sticky_Directories In a standard System V.3 UNIX system, a user has the ability to delete a file if (and only if) the user has write permission on the directory that contains the file. There are some directories in a system (for example, /tmp and /usr/tmp) that need to be writable by all users on a given system. With the standard System V.3 permission rules, this gives any user the ability to delete files belonging to other users. For example, any user on the system could issue the command rm /tmp/*, deleting everything in /tmp. In order to provide the ability to prevent this, Release 4D1-3.3 IRIX provides a new feature adopted from other variants of UNIX: turning on the sticky mode bit on a directory changes the rules for deleting files from that directory. A user can delete a file in a sticky directory if and only if the user has write permission on the directory and at least one of the following is true: - the user is the superuser - the user owns the directory - the user owns the file to be deleted - the user has write permission on the file to be deleted Set the sticky bit with the command: chmod +t /tmp A directory with the sticky bit set can be recognized from the output of ls -l by the t in the world execute permission position of the modes field: drwxrwxrwt 3 sys sys 512 May 10 20:46 /tmp Note that if the sticky bit is set on a directory, the permission checking rule given above applies to any operation that would result in a file being removed, not just a rm(1) command. If /tmp is a sticky directory, then the command mv foo /tmp/bar will fail unless the executing user has the ability to delete the file /tmp/bar according to the rule given above. Every time the system is booted into multiuser mode, the script /etc/rc2.d/S58RMTMPFILES is executed. This script deletes everything in the /tmp directory and creates the directory /usr/tmp if it doesn't already exist, among other functions. In Release 4D1-3.3, this script has been changed to set the sticky bits on /tmp and /usr/tmp to enable the more restrictive file deletion semantics described above. If you prefer the previous (less restrictive) behavior, modify the script by finding the line chmod 1777 /tmp /usr/tmp and changing it so that it does not turn on the sticky bit: chmod 777 /tmp /usr/tmp 4.1.3 New_Commands_and_Utilities o The new passmgmt(1M) command allows the superuser to update /etc/passwd entries. o The new unifdef(1) command automatically strips or reduces #ifdef constructs in C source or any source filtered by the C preprocessor, cpp(1). o Keyboard auto-repeat rate may be adjusted on workstations using the new keywarp(6) command. o The new -A option to ls(1) lists entries beginning with a dot except for the . and .. directories. o The new -s option to hostname(1) trims any domain information from the printed name. o The new -f option to which(1) uses the current path by ignoring the .cshrc file. o login(1) has several new options to require passwords, record and display last login times, and syslog(3) failed and successful network or tty login attempts. See the login(1) manual entry for details. 4.1.4 SCSI_Driver There is a new feature in the SCSI driver that can be used when devices must be connected that either do not support the disconnect feature or do not work correctly when it is enabled. This feature may be enabled by editing the file /usr/sysgen/master.d/scsi and changing the value of the variable scsi_enable_disconnect from 1 to 0. Enabling this feature (setting the variable to 0) may impact performance somewhat when multiple devices are on the SCSI bus, since operations on different devices will no longer be overlapped. Also note that when this feature is enabled, some Exabyte tape drives may not work correctly unless switch #6 is set. 4.2 Additions_to_the_Graphics_Library o A new routine, frontface, controls the removal of frontfacing polygons. Its function is similar to backface, except that it discards polygons whose vertices are in the opposite order. See the frontface(3G) manual entry for more information. o lmcolor is now functional on IRIS-4D B and G models. 4.2.1 PowerVision_Commands The 4D1-3.3 release introduces PowerVision, a new high-performance graphics architecture. Many new GL commands have been added and several existing routines have additional functionality on PowerVision systems. 4.2.1.1 New_Routines The following list describes the new Graphics Library routines available for use with PowerVision. These routines are explained in more detail in the Graphics Library Programmer's Guide and the Graphics Library Reference Manual. acbuf operates on the accumulation buffer acsize specifies the number of bitplanes per color component in the accumulation buffer afunction specifies alpha test function bgnqstrip begins a quadrilateral strip vertexes clipplane specifies a plane against which all geometry is clipped endqstrip ends a quadrilateral fogvertex specifies fog density for per-vertex atmospheric effects getscrbox reads back the current computed screen bounding box mswapbuffers swaps multiple framebuffers simultaneously nmode specifies renormalization of normals pixmode specifies pixel transfer mode parameters polymode controls the rendering of polygons polysmooth specifies antialiasing of polygons sclear clears the stencil bitplanes to a specified value scrbox controls the computation of a screen space bounding box scrsubdivide subdivides lines and polygons to a screen- space limit stencil alters the operating parameters of the stencil stensize specifies the number of bitplanes to be used as stencil planes swritemask specifies which stencil bitplanes can be written t specifies a texture coordinate tevbind selects a texture previously defined with texdef(2D) to be used for current drawing tevdef defines a texture mapping environment texbind attaches a predefined texture to one of several system textures texdef2d converts a 2-dimensional image into a texture texgen specifies automatic generation of texture coordinates 4.2.1.2 Changes_to_Existing_Routines o The overlay and underlay routines has additional functionality under PowerVision. In earlier GL releases, calling overlay with a non-zero argument forced underlay to zero. Because PowerVision supports simultaneous overlay and underlay, it simply allocates resources based on the sequence of overlay, underlay, singlebuffer, and doublebuffer calls, giving priority based on the calling order (specifically, the last call wins). The IRIS PowerVision supports allocation of 8 bitplanes, either single or double buffered, to overlay or underlay. When this allocation is requested, the alpha bitplanes (which must be present) are used, and are therefore not available for storage of destination alpha values (see blendfunction). Single or double buffer operation is specified by calling either singlebuffer or doublebuffer while in drawmode OVERDRAW or UNDERDRAW, prior to calling gconfig. Currently only colormap mode is available in the PowerVision overlay and underlay framebuffers. Double buffered overlay and underlay are swapped when mswapbuffers is called with the OVERDRAW or UNDERDRAW bit set. o drawmode The values of the drawmode tokens have changed. Old values are still accepted by drawmode for compatibility. getdrawmode() will return the new values. Programs should always use the symbolic constants so they don't have to depend on the values of these constants. o mmode A new matrix mode parameter value, MTEXTURE, has been added to mmode for texture mapping on PowerVision systems. See mmode(3G) for information on the use of MTEXTURE. The texture matrix is also initialized to the identity matrix. o readsource is always SRC_AUTO when not in NORMALDRAW drawmode. 4.2.1.3 PowerVision_Compatibility_with_Other_Systems This section lists various compatibility issues between PowerVision (VGX) and GT/GTX systems. Several Graphics Library routines exhibit different functionality on these two types of systems. o subpixel By default subpixel mode is FALSE. GT/GTX systems achieve their rated performance in this default mode. On VGX systems, however, performance is substantially degraded when subpixel is FALSE. The published performance figures for VGX machines are valid only when subpixel mode is explicitly set to TRUE. Because the scan conversion of lines is affected by subpixel mode, some 2-D orthographic programs may require subpixel FALSE while drawing lines. Refer to the subpixel man page for details. o move/draw Lines drawn using the move/draw sequence are forced to be flatshaded on the G and GT/GTX models. However, both the Personal Iris and the VGX systems shade lines according to the shademodel argument. o rmv/rdr Matrix operations are not supported between rmv/rdr, move/draw sequences. o zwritemask The argument to zwritemask has always been documented to be a mask, however the GT/GTX and Personal Iris interpreted it as a flag. The VGX interprets the argument as a true mask. Therefore, zwritemask(TRUE) will mask the zbuffer to one bit on VGX systems. o old style polygons/circf/arcf By default, the GL renders filled circles and arcs as old styled polygons. Currently, on the VGX systems, the filling algorithm of these polygons has problems with missing pixels. To work around this, one can call glcompat(GLC_OLDPOLYGON, 0) prior to drawing filled circles, arcs and old styled polygon drawing commands. However, using old styled polygon commands to draw screen aligned rectangles works perfect. Therefore, any application that uses this feature will continue to works. An example of this is the window manager 4sight. o rot accuracy The rot routine on the VGX systems is more accurate than the GT/GTX systems. o gversion Some programs that look for specific return values from gversion quit if they don't see the GT/GTX values. Such programs fail to work on VGX machines. In general, use of getgdesc to query for specific capabilities should replace all uses of gversion because this behavior will be forward-compatible. o concave decomposition On GT/GTX systems, concave decomposition is always on. On Personal IRIS and PowerVision systems, however, concave decomposition is off by default. You must issue a concave(TRUE) call to instruct the system to expect concave polygons. See the concave(3G) manual entry for more information. 4.2.1.4 Known_User_Errors This section lists various common user errors that we have found. o lmbind/MSINGLE lmbind() can not be used in the MSINGLE mode. The lmbind manual page notes this. However, we did find some applications that did this. o zbuffer/czclear On the GT/GTX machines, the zbuffer value is represented as a 23 bit unsigned integer. This meant that the getgdesc of GD_ZMIN is 0. The GD_ZMIN value is typically used for lsetdepth() and czclear(). However, some applications used czclear with z hardcoded 0 for the z value. This will break on the VGX and Personal Iris machines as GD_ZMIN is 0xff800000, or a large negative value. 4.2.2 New_Display_List_Commands The Graphics Library included with the 4D1-3.3 release adds display list support for high-performance graphics commands previously restricted to immediate execution. This capability was achieved without a significant impact on immediate-mode performance. The Graphics Library's implementation of graphical objects (see Chapter 16 of the Graphics Library Programmer's Guide) calls for GL drawing primitives to behave differently depending on whether the GL is in immediate-mode or object- creation mode at the time a primitive is invoked. Previous implementations of the GL included conditional branches within each GL routine that selected between immediate- drawing and object-creation behaviors. This implementation precluded the use of the high-performance drawing routines in graphical objects because the requisite conditional branching reduced performance. Beginning with Release 4D1-3.3, the selection between immediate-mode and object-creation behaviors is implemented by switching shared-library jump tables when entering and leaving object-creation mode. This implementation allows the high-performance drawing routines to perform well enough that they can be used within graphical objects. The following routines are now available within graphical objects: bgnpoint, endpoint bgnline, endline bgnclosedline, endclosedline bgnpolygon, endpolygon bgntmesh, endtmesh, swaptmesh n3f c3s, c3i, c3f c4s, c4i, c4f v2s, v2i, v2f, v2d v3s, v3i, v3f, v3d v4s, v4i, v4f, v4d cpack wmpack czclear linesmooth, pntsmooth, subpixel In earlier releases, executing these routines in object- creation mode was strictly prohibited by the manual. If a program called them, they did immediate drawing, as if the GL were in immediate mode. Beginning with Release 4D1-3.3, calling these routines in object-creation mode results only in building the current object; no immediate-mode drawing is done. The ability to use these routines within graphical objects comes at the price of slightly lower performance for the routines that switch the GL into and out of object-creation mode: makeobj, closeobj, and editobj. This implementation depends on the use of shared-library jump tables, which are not available to the non-shared version of the GL. In order to make the non-shared version function identically to the shared-library version, its versions of the high-performance drawing routines must use conditional branching, and they are thus somewhat slower than they were before. Use of the shared GL is strongly encouraged for a number of reasons; this is yet another reason to use it instead of the older, non-shared version. To use the shared-library version, use -lgl_s instead of -lgl when compiling. 4.2.3 NURBS The NURBS interface to the Graphics Library includes these enhancements in Release 4D1-3.3: o 8th-order curve rendering on the Personal IRIS o unlighted rendering on all IRIS-4D platforms o new surface capabilities In previous releases, users could describe only the geometry of a trimmed NURBS surface and render it. With this release they may also specify texture coordinates and color coordinates in addition to positional coordinates. This information is communicated through the use of the type field in the call to nurbssurface. Specifying texture and color surface information has an effect only on VGX machines. On IRIS-4D G, GT, GTX and Personal Iris, such specifications have no effect. 4.2.4 Graphics_Input_Queue A new GL subroutine qgetfd(3G) allows a GL program to use the select(2) system call to determine when there are events waiting to be read in the graphics input queue. A call to qgetfd returns a file descriptor that may be used as part of the readfds parameter of the select system call. When select indicates that the file descriptor associated with the graphics input queue is ready for reading, a call to qread(3G) or blkqread(3G) will not cause the program to block. See the description of select in the IRIX Programmer's Reference Manual for more information. A new GL subroutine qcontrol(3G) provides an interface for various administrative functions in the graphics input sub- system. With qcontrol, it is possible to add new types of input devices and programmatically control the keyboard auto-repeat rate and mouse acceleration parameters. 4.2.5 Multiple_Screens As of this release, the Graphics Library supports applications opening windows on more than one screen. The routines added to provide this support are: o scrnselect selects the screen upon which windows are opened. o scrnattach attaches the input focus to a specified screen. o getwscrn returns the screen upon which the current window is displayed. For more information, see the 4Sight Programmer's Guide and the manual entries for the individual routines. 4.3 Additions_to_Demos Release 4D1-3.3 includes these new demos: ___________________________________________________________________ Demo Description ___________________________________________________________________ trimnurbs trimmed NURB (non-uniform rational B-spline) surfaces revolve interactive surface of revolution editor ___________________________________________________________________ ___________________________________________/No man is a failure who has friends Trevor Paquette ICBM:51'03"N/114'05"W|I accept the challange, body and soul, {ubc-cs,utai,alberta}!calgary!paquette|to seek the knowledge of the ones of old paquette@cpsc.ucalgary.ca | - engraved on the Kersa Blade of Esalon   Received: from vmb.brl.mil by VMB.BRL.MIL id ab01973; 25 Jun 90 23:34 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa01898; 25 Jun 90 23:23 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01862; 25 Jun 90 23:13 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa11535; 25 Jun 90 22:45 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA18864; Mon, 25 Jun 90 19:16:56 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 01:10:48 GMT From: Trevor Paquette Organization: The University of Calgary Subject: Irix 3.3 Bug_Fixes Part 1/2 Message-Id: <1990Jun26.011048.24806@calgary.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL ------------------------- Irix 3.3 Release Notes 6. Bug_Fixes Taken from the command 'pcat /usr/relnotes/standard/ch6.z' and edited to remove underlining etc.. ------------------------- 6. Bug_Fixes This chapter describes bug fixes to: o IRIX o program development tools o software installation tools o networking o the Distributed Graphics Library (DGL) o mail programs o demos o 4Dgifts o graphics o 4Sight o IRIS WorkSpace o the X Window System A Silicon Graphics software change request (``SCR'') number appears after many of the bug fixes in this chapter. 6.1 Bug_Fixes_to_IRIX o Missing arguments to a debugging printf() and return statements in random(3) have been added. o The lpshut(1M) command would occasionally fail to completely shut down the lp system when the system was printing via the network. lpshut has been fixed to ensure that the lp system is actually stopped before it finishes. (SCR 7628) o The new syslogd daemon is provided to log kernel messages (e.g. disk errors) to the disk. (SCR 3079) o The /etc/shutdown continuation query message has been clarified. (SCR 3118) o vi tags now generate search forward commands, so the initial tag search will work. Moving below the match and issuing a :ta command again will fail (assuming there are no more matches). (SCR 3265) o The /usr/people/guest/.profile and /.profile files no longer set the $PATH. The ls function and tset invocation in these profiles is refined. (SCR 3279) o The fs(4) manual entry now consistently uses the term ``fs'', not ``efs''. (SCR 3319) o The continue statement in awk no longer occasionally invokes the next statement. (SCR 3493) o The join(1) command no longer core dumps if the number of fields in the input line is less than the join field specified by -j. (SCR 3604) o The -prune option has been added to the find(1) command. (SCR 3790) o awk(1) now has an improved usage error message. (SCR 3925) o mv(1) now maintains file ownership and times when moving files across file systems. (SCR 3941) o The graphics input queue now allows dynamic size allocation via lboot(1M). (SCR 3942) o exec(2) now allows tuning of NCARGS argument list size. (SCR 4145) o Several dozen words have been added to the spell dictionary of known words. (SCR 4364) o The bc(1) command no longer dumps core on the di command. (SCR 4502) o dircmp and file now recognize dangling symlinks. (SCR 4699) o odump is now documented in the Programmer's Guide as an object file dump command. (SCR 4718) o now incorporates portions of the now obsolete . (SCR 4876) o now incorporates portions of the now obsolete . (SCR 4883) o The nawk implementation limits have been documented in the Programmer's Guide. (SCR 5014) o Chapter 13, ``Shared Libraries'' of the Programmer's Guide has been rewritten to describe Silicon Graphics's version of the shared libraries (rather than AT&T's version). (SCR 5015) o The PATH environment variable is no longer set in any of the default provided shell scripts (such as various .profile, .cshrc, and .login scripts). A default PATH is set by each program that establishes a login session, such as /bin/login, /etc/gl/pandora. (SCR 5071) o Configurable security features have been added to login. Logging of successful and unsuccessful login attempts can be enabled with the syslog keyword in the configuration file /etc/config/login.options. The number of attempts permitted and the disable time after unsuccessful attempts can also be configured in /etc/config/login.options. (SCR 5073) o The ls -F option and special symbols /, * and @ are now documented in the ls(1) manual entry. (SCR 5074) o tic now allows longer filenames. (SCR 5232) o vi now handles SIGWINCH to change window size. (SCR 5324) o The whatis(1) BSD command is provided in Release 4D1- 3.3. (SCR 5370) o The w(1) BSD command is provided in Release 4D1-3.3. (SCR 5398) o no longer requires to be explicitly included first. (SCR 5534) o The expr(1) manual entry now documents the match operator. (SCR 5682) o The explanation of the getitimer(2) return value has been clarified in the manual entry. (SCR 5726) o setvaluator now always generates events properly. (SCR 5745) o The hinv command reports CPU and FPU revisions, replacing the obsolete showconfig kernel routine. (SCR 5755) o prctl has been enhanced to allow the sending of a signal to an entire share group upon exit of any member of the group. This allows sproc child processes to be signaled if a parent process dies. (SCR 5759) o The sadp command has been deleted since it cannot work with intelligent disk controllers. (SCR 5802) o The /etc/killall executable is marked set group id == sys to allow it to access /dev/kmem on behalf of any user. (SCR 5808) o The IP4 status LEDs now show 1 second ``heartbeats''. (SCR 5810) o The argument type description in the signal(2) manual entry has been corrected. (SCR 5908) o The descriptions of the h and l character positioning commands have been corrected in the vi(1) manual entry. (SCR 5947) o The information on FILES in the math(5) manual entry has been corrected. (SCR 6015) o pty: adjusted STREAMS pseudo-terminal driver to avoid resource deadlocks. (SCR 6023) o The fclose(3S) library routine now avoids forcing errno to ENOENT. (SCR 6036) o now declares sys_errlist[] and sys_nerr. (SCR 6054) o tar now provides for partial ability to handle multiple volume backups onto QIC-02 drives. (SCR 6090) o ls(1) manual entry now documents the -L and -H options. (SCR 6129) o find(1): The command find . -name '+*' -print now finds those files whose name begins with '+', instead of finding all files. (SCR 6174) o The /tmp directory is no longer removed during the boot process if it is a symbolic link or mount point. (SCR 6210) o tar(1) has been fixed so that "tar t " lists only requested filenames. In previous releases, this command listed all names in the tar archive. (SCR 6264) o The rint(3M) math routine is provided to round in direction of rounding mode. (SCR 6370) o script(1) now passes control-Z and control-C to its subshells. (SCR 6398) o tail(1) now buffers up to 256 kilobytes of data. (SCR 6404) o The sh(1) internal command ulimit has been fixed to display a value of zero (0) correctly. (SCR 6408) o memory: Release 4D1-3.3 provides improved management of process memory and core file needs. The default core file size is now adjustable. Also, there are no more out of swap messages printed. You can adjust the maximum RSS size of a process (thus guaranteeing that some resources are always available) and the maximum stack size. The allocation of swap space is now delayed until it is actually needed. This allows a process to malloc very large, infrequently-used data areas to have a virtual memory size that is larger than available swap space, as long as the pages of memory that are actually touched fit in swap and memory. When swap finally runs out, the ``greediest'' process is selected by the kernel and killed, allowing other processes to continue correctly. (SCR 6416) o The new command ident(1) searches files for RCS keywords. (SCR 6480) o The curses library libcurses.a has been converted to use ioctl(FIONREAD) and select(2) to improve handling of input. (SCR 6499) o The BSD system calls getrlimit(2) and setrlimit(2) have been added to IRIX in this release. The csh(1) builtin commands for manipulating resource limits are also provided. (SCR 6515) o Obsolete kernel fastcallout table removed from /usr/sysgen/master.d/kernel. (SCR 6539) o The killall(1M) manual entry describes handling of numeric process names. (SCR 6541) o The reference to mntent.h in the setmntent(3) manual entry is now correct. (SCR 6543) o awk and oawk now emit correct line numbers in error messages for awk scripts with comments. (SCR 6549) o The mknod(1M) manual entry now identifies the correct location of list of major device numbers. (SCR 6573) o vi(1) no longer creates temporary files that are several times the size of the file being edited. vi(1) also now gives more meaningful error message when the system is out of disk space. (SCR 6584) o Directory access times are now changed if the directory is read, for example, by the ls(1) command or the scandir(3C) library routine. (SCR 6615) o ps(1): The output of the -l option now provides a two byte F field and wider SZ and TIME fields. (SCR 6633) o The RSS output field from the -l option of ps(1) is now based on kernel data that is updated every second and the calculation of this value is improved to handle shared regions and mapped physical device memory, as described in the ps(1) manual entry. (SCR 6663) o Declarations for the routines utmpname, setutent and endutent have been added to the include file . (SCR 6667) o vi(1) no longer creates zombie processes when using the ! command to filter a file's contents. (SCR 6727) o The kernel's ability to create a crash file in /usr/adm/crash in the unfortunate event of a panic is more reliable. (SCR 6733) o The Mail(1) command now accurately detects the size of the window it is invoked in. (SCR 6745) o stdio.h: irrelevant portions for VAX and MIPS hardware have been removed. (SCR 6750) o The gr_osview -N option now works with remote host accounts that have password protection, as long as the related .rhost file allows access. (SCR 6814) o The new setrlimit(2) system call provides a way to limit core file sizes. (SCR 6887) o The -M and -d options to the RCS command co have been fixed to calculate time correctly. Also, a possible core dump condition is eliminated. (SCR 6909) o The dialwarp command is now documented in a manual entry. (SCR 6974) o setitimer now allows for better timer resolution (less than 1 millesecond) using ftimer. (SCR 6976) o at(1) now uses the correct group ID (gid) to run a task in cases where the user previously changed groups using newgrp(1). (SCR 6993) o The terminfo item sgr0 is now specified in the terminfo file for iris-ansi TERM type. This allows the request attrset(A_NORMAL) to work inside programs, even when following an attrset(A_BOLD) request. (SCR 7006) o The Berkeley BSDsetpgrp(2) and vhangup(2) routines now have manual entries. (SCR 7011) o tpsc(7M): The tape driver is enhanced to allow writing data after the EOT marker, as required by ANSI X3.27 standards. (SCR 7054) o /usr/include/gl/dials.h: Removed bogus comment. (SCR 7073) o memory allocation: Routines that allocate stack space, such as the GNU routine alloca, can now pass the address of that space to the kernel, and expect that the stack will be grown appropriately if needed to create the space. Previously, system calls requiring such space could fail with errors such as EFAULT, after the kernel had failed to automatically grow the required space. (SCR 7080) o make(1): Added a description of the $(@D) and $(@F) macros to the make(1) manual entry. (SCR 7106) o make(1): The macro $(@F: .o= .c) now works inside a suffix rule. (SCR 7145) o m_fork(3P) no longer produces a segmentation violation. (SCR 7159) o References to the obsolete initstate 5 have been deleted from the shutdown(1M) manual entry. (SCR 7174) o bstream(1): The undocumented -B option is no longer available. A bug that could cause bstream to not write a final, partial sized block to tape has been fixed. A bug that could cause bstream to hang has been fixed. (SCR 7301) o The make(1) manual entry now documents VPATH. (SCR 7334) o console baud rate: the dbaud parameter in the bootrom command monitor now changes the graphics console baud rate. (SCR 7340) o The dbg(4) manual entry is now also available using the name debug, for convenience of access. (SCR 7363) o virtual memory: The performance of the virtual memory paging daemon is improved to minimize time spent computing page reference counts. (SCR 7367) o The awk(1) command now queries the kernel to determine the available number of open file descriptors, instead of allowing just 15 open files. (SCR 7391) o mkdir(1): The -p option now exits with a zero (0) status if it is successful. (SCR 7394) o The usconfig(3P) routine now returns the documented values. (SCR 7407) o The mkfile(1M) command is now described in a manual entry. (SCR 7408) o syslogd(1M): The obsolete -m option to syslogd is no longer mentioned in the manual entry. (SCR 7458) o The System V semaphore and message IPC structures are now initialized to zeros before being allowed to be accessed from applications with semget() or msgget(). (SCR 7459) o The include file /usr/include/sys/invent.h is now C++ compatible. (SCR 7466) o The imon(7M) command is now documented in a manual entry. (SCR 7467) o The halt(1M) manual entry has been rewritten for improved clarity. (SCR 7488) o Various virtual memory bugs allowing system hangs or instability in the virtual memory paging system have been fixed. (SCR 7490) o The bootp(1M) manual entry no longer describes the obsolete -h option. (SCR 7514) o The setsym(1M) command is now described in a manual entry. (SCR 7574) o Password aging is now described in the passwd(4) manual entry. (SCR 7613) o sed(1) no longer dumps core on an empty expression, such as 's///'. (SCR 7637) o File creation performance has been improved. Creation still writes to the disk synchronously to increase file system robustness, but it doesn't take as long. (SCR 7640) o The vi(1) editor now recognizes changes in the size of the window it is invoked in. (SCR 7646) o A bug in regcmp(3X) that could cause a core dump has been fixed. This bug could be seen by invoking the command man *. (SCR 7700) o swap(1M): Fixed various bugs that could cause swap -d (delete swap space) to fail or to hang the system. (SCR 7704) o libsun.a: The strdup() routine, which is available in libc.a2, has been removed from libsun.a. (SCR 7729) o The exit(2) manual entry now correctly describes exit's effect on open files and its interactions with sproc(2). (SCR 7747) o The nawk(1) command no longer core dumps on null bodied functions. (SCR 7772) o The file /usr/include/dirent.h now correctly defines MAXNAMLEN to be 255. (SCR 7775) o A bug in gr_osview(1) that could cause gr_osview to hang the system has been fixed. (SCR 7782) o The following routines have been added to libds.a: read08(), reservunit16() releaseunit17(), These routines were already documented in the dslib(3) manual entry. (SCR 7800) o The pcreate(3C) routine is now cross-referenced in the manual entries for fork(2), exec(2), prctl(2), sproc(2) and system(3C). (SCR 7824) o The descriptions of the hinv(1M) options have been corrected in the manual entry. (SCR 7826) o The vi(1) command B now scrolls backward two lines less than a full screen, instead of two lines more. (SCR 7830) o The directory /usr/bin/X11 has been added to the search path of whereis(1). (SCR 7842) o The default PATH for the tutor login has been fixed. (SCR 7852) o Bugs in the more(1) command (/usr/bsd/more) that could cause more to leave the terminal in a raw state after an interrupt have been fixed. (SCR 7854) o The logname(1) manual entry has been changed to correctly describe its use of cuserid(3S). (SCR 7875) o The login(1) command has been changed to allow for $HOME directory pathnames that are up to PATH_MAX (which is 1024) characters long, instead of only 64 characters. (SCR 7901) o savecore(1): System crash files in located in /usr/adm/crash are now created with file modes of 0600. This close as security loophole. (SCR 7904) o The sar -b option now checks for zero and negative results to avoid displaying nonsensical values. (SCR 7906) o The man(1) manual entry now documents the PAGER option. (SCR 7925) o exec(2) no longer fails with the error message Insufficient memory to allocate 8 pages when additional free swap space is still available but main memory is all in use. (SCR 7936) o The subj(1) command now works when invoked from csh. (SCR 7951) o The inittab(4) manual entry now states that the ID field in /etc/inittab can have up to four characters. (SCR 7956) o The sar (1) manual entry now describes all of the columns output by the -t option. (SCR 7974) o The csh(1) manual entry now documents the child environment variable. (SCR 7975) o The SHACCT shell script accounting feature is now available for sh(1). (SCR 7981) o The mt(1) manual entry now describes where to find the values for the Status field output by the mt status command. (SCR 8006) o more(1) - Reliably cleans up the --More-- prompt at completion. (SCR 8018) o getty(1M) - The getty manual entry documents the -s option, to allow to getty's on the same line to avoid colliding. (SCR 8025) o mknod(1M) - The mknod manual entry now refers to the correct file for major device number assignment. (SCR 8031) o cron(1) - The cron command now sets the USER environment variable when running commands. (SCR 8032) o intro(2) - The introduction to section two of the manual now describes EAGAIN, O_NDELAY, empty streams and locked files. (SCR 8050) o cpio(1) - The cpio manual entry describes the -H and -L options. (SCR 8100) o ctags(1) - The ctags manual entry describes the -v option. (SCR 8126) o sysfs(2) - The sysfs manual entry describes its expected fsname argument. (SCR 8134) o exec(2) - The exec manual entry now refers to ARG_MAX instead of to the obsolete explicit value of 5120. The current value of ARG_MAX is 10240. (SCR 8168) o which(1) - The -f option is provided, to suppress reading of ~/.cshrc. (SCR 8201) o apropos(1) - The BSD Unix commands apropos and makewhatis are provided. (SCR 8312) o system(3S) - Thanks to improvements in the virtual memory management of IRIX, the system(3S) library routine can now efficiently handle the invocation of a command from within a large process, without requiring enough swap space to store a temporary additional copy of the large process. (SCR 8366) o exec(2) - The exec manual entry now discusses share group processes. (SCR 8367) o vi(1) - The vi editor has improved handling of the numeric keypad. (SCR 8474) o crontab(1) - The crontab command no longer allows null length files from standard input, to avoid allowing the user to accidentally empty their crontab file. (SCR 8522) o passwd(4) - The valid values for user id (uid) and group id (id) are extended from 29999 to 60000. (SCR 8597) o env(1) - The env command no longer has a limit on the number of environmental variables handled. It had been limited to 100 variables. (SCR 8625) o closedir(3B) - The closedir library routine now works with other malloc and free library routines even when the free routine clears or alters memory while freeing it. Before, the closedir routine could fail to close the file descriptor in use. (SCR 8633) o more(1) - The more command no longer is confused by requests to skip more pages than remain to be displayed. (SCR 8636) o login(1) - The login command now supports password aging and limits on the number of successive failed attempts. (SCR 8640) o diff(1) - The diff command is upgraded to the Berkeley version, and supports the -c context option. (SCR 8653) o login(1) - Fixed a problem that prevented logging in with graphics when certain other jobs were executing in the background. (SCR 8675) o sysconf(2) - The number of characters allowed in the exec(2) argument list is now configurable. (SCR 8677) o acctmerg(1M) - The acctmerg can now merge up to 100 (see getdtablesize(2)) files at once, instead of just 10 files. (SCR 8726) o shutdown(1M) - The /etc/shutdown script no longer runs /etc/rc0 to terminate multi-user processes, but rather lets init run /etc/rc0 after single user state is entered. This solves a problem that had prevented entering single user mode while using pandora(1). (SCR 8790) o rmt(1M) - The remote tape facility is enhanced to provide greater compatibility with other Berkeley Unix based systems and with older IRIX releases. (SCR 8865) o mousewarp(6D), keywarp(6D) and dialwarp(6D) - These facilities are added to allow user configuration of input device response speeds. (SCR 8866) o csh(1) - The foreach construct now allows an index variable name that has embedded digits. (SCR 8880) o sh(1) and csh(1) - The shells now have builtin commands to allow setting limits on coredumpsize (the size of the largest core dump that can be created), memorysize (the maximum amount of physical memory allocated to a single process), and other memory and file limits. By setting a reduced coredumpsize, users can avoid having their system take minutes to dump a large application's core. (SCR 8911) o spell(1) - The spell command no longer fails in the event that the invoking environment happened to have a variable set by the name of "D". (SCR 9012) o inittab(4) - The inittab manual entry now describes the (limited) comment conventions of the inittab file. (SCR 9022) o man(1) - The man command now supports the -k (keyword) option. (SCR 9032) o multgrps(1) - The Berkeley multiple groups feature is supported. (SCR 9037) o prtvtoc(1M) - The prtvtoc now displays the correct number of bytes per sector, even on disks formatted by old versions of the fx(1M) command. (SCR 9100) o documentation - A caveat is included in the Release Notice, providing additional instructions for section 9.2 of the "Writing Device Drivers for Silicon Graphics Computers" guide. These instructions describe how to test a terminal before selecting "setenv console d" in the boot rom monitor. (SCR 9128) o IRIX - A virtual memory bug in IRIX is fixed, which had caused systems to occassionally hang after many hours of heavy swapping. (SCR 9133) o su(1M) - The su command now exits with an appropriate message if it is unable to open /dev/tty to obtain a password. (SCR 9137) o mkdir(2) - Newly created directory blocks are now initialized to zero's. (SCR 9183) o sh(1) and csh (1): The ability to use setuid and setgid permissions on shell scripts is restricted. This ability is enabled by a kernel configuration option. By default this option is disabled. See the variable nosuidshells in the file /usr/sysgen/master.d/kernel. The setuid shell script /etc/gl/startconsole is rewritten as a C compiled program, so that the setuid shell script feature is no longer required by any Silicon Graphics released software. These changes improve system security. (SCR 9287) o dcopy - References to this unsupported command are removed from the System Administrator's Guide. (SCR 9340) o script(1), vi(1) and more(1) - The vi and more commands now work while using the script command to capture terminal sessions. (SCR 9363) 6.2 Bug_Fixes_to_Program_Development_Tools 6.2.1 C_Compiler_Bug_Fixes o cc now generates a warning when -#, an unsupported option, is used. (SCR 5308) o The fatal internal error schain botch is no longer caused by any source program. (SCR 6106) o << and >> in constants now get the right signed or unsigned type. (SCR 5681) o Previously, ++ was ignored in complex circumstances. It is no longer ignored. (SCR 5478) o enum declarations in inner scopes are now handled correctly. (SCR 6145) o Prototypes are now handled correctly. (SCR 4678, 4895, 5088, 5904, 6153, 6242, 6462) o Coredumps on illegal programs were fixed. (SCR 4351, 5378, 5458, 7168, 7252) o volatile now generates the correct types and is accepted wherever it is allowed in ANSI C. (SCR 5454, 6378, 6381) o Floating point expressions (d > d1) did not generate an integral 0/1 result. (SCR 7013, 7223) o void and void* now work. (SCR 3193, 4797, 6010, 6256, 6773, 7285) o Register declarations sometimes used to cause a ugen coredump or error message. (SCR 5263, 5716) o The ternary operator ?: with float types used to cause as1 coredumps. (SCR 6775, 7191, 8484) o enum values are now treated as another name for int. (SCR 7136, 8609) o Some struct pointer function prototypes used to cause a ccom coredump. (SCR 7321) o The const keyword is now understood. (SCR 6257) o Structure returns no longer cause argument passing problems. (SCR 8853) o Using float to and from unsigned conversions now works correctly. (SCR 5441) o Several optimizer (cc -O2) bugs were fixed. (SCR 4414, 4585, 4624, 4918, 5084, 5351, 5442, 5455, 5577, 6099, 6146, 6371, 7088, 7236, 8581) 6.2.2 C_Preprocessor o Escaped newlines in strings resulted in wrong line numbers being assigned. (SCR 6219) 6.2.3 Other_Development_Tools o The assembler would coredump if the user specified an assembly-time divide by zero. (SCR 6981) o The reorganizer (assembler) would generate illegal reorganizations in certain rare circumstances. (SCR 7332) o cflow now generates type output correctly. (SCR 7283) o cxref had several problems. (SCR 6376, 6380, 6386, 6950, 7357, 8596) o ar now uses the TMPDIR environment variable to determine the location of temporary files. (SCR 5387) o ar used to coredump with certain bad object files as input. (SCR 6041, 6105) o nm used to print out erroneous symbol data under specific circumstances. (SCR 5475) o cb now has improved option checking so that improper options will not silently put it into reading standard input. (SCR 5307) o lint now reports unused variables. (SCR 5374) o ldexp(3C) now returns 0.0 if its first argument is 0.0 instead of treating 0.0 as a denormalized number. o Several bugs in make(1) have been fixed: - the -p option now doesn't print extraneous things; all output goes to stdout; when -d is on, stderr/stdout are flushed so output isn't interspersed. (SCR 6086) - the -u option was added to unconditionally execute a command (SCR 6196) - The manual entry now documents the $(@D) and $(@F) macros (SCR 7106) and VPATH (SCR 7334). - the macro $(@F: .o= .c) now works inside a suffix rule. (SCR 7145) o Several bugs in pmake(1) have been fixed: - a missing rm was added to the .sh.out rule in /usr/include/make/system.mk - the incorrect transformation rules for archives were removed - pmake now properly checks lines beginning with include properly to decide if the line is a target or include command 6.3 Bug_Fixes_to_Software_Installation_Tools o A message will be displayed while the system is coming up if there are configuration files that have changed status during a recent software installation. (SCR 5045) o Annotation explains the meaning of the .O and .N suffixes in the versions config and versions changed output. (SCR 5051) o The versions user command will no longer descend into NFS-mounted directories during its scan of the file system. (SCR 5767) o The miniroot will now display the current system date and time before the installation tool runs, as a precaution against installing software with an incorrect system date. Resetting, if necessary, can be done through a shell escape. (SCR 5805) o The installation tool would sometimes retain an incorrect notion of the physical position of the tape when other errors occured. This has been corrected. (SCR 5869) o The notification of which subsystems cannot be installed on the currently-running IRIX system was not being given at an appropriate time. It has been integrated with the other pre-installation checks. (SCR 5881) o During deletion of installed products, the tracking of subsystems that had changed names or other grouping was not always done correctly. Subsystem tracking for purposes of removal has been fixed. (SCR 5896) o The versions command has been built into inst, and can be run at any time as a menu item, as well as a separate IRIX command. (SCR 5922) o The default root directory was not being set correctly for versions commands executed from a chrooted shell. This caused unexpected versions behavior. (SCR 6068) o Some disk full conditions were not properly detected by previous versions of inst. Writes to a full disk are now detected, and generate the appropriate error conditions. (SCR 6069, 6070, 7491) o There were several weaknesses in the file system mounting, unmounting, and making logic in inst, sometimes causing incorrect behavior. These have been corrected. (SCR 6252) o Many reliability issues in the handling of the online installation versions history have been addressed. These improvements solve certain problems where subsystems and files were not being removed properly. (SCR 6253) o The frequency of disk space recalculations has been greatly reduced. Typically, the disk space calculations are done only once per software distribution access. (SCR 6396, 7554) o The installation tools now retain a record of all products that are available during the installation process, even those that are not installed. This information tends to improve the value of the default subsystem selections in future installations. (SCR 6410) o The error message for the insufficient disk space condition has been improved. Previously, the name of the offending file system would sometimes not appear. (SCR 6558, 6648, 9025) o While running in the miniroot, older versions of inst would temporarily convert symbolic links to directories from absolute to relative form, so that they would resolve during installation. The path interpretation is now done without altering the link. (SCR 6689, 8128) o Network timeouts during software installation could sometimes result in a core dump. This has been corrected. (SCR 6695) o Inst no longer attempts to access a remote distribution server as root; the default is guest. (SCR 6709) o The miniroot no longer uses default /etc/sys_id and /etc/hosts configuration files. These files are derived at the first attempt to access the network. (SCR 7128) o The reliability of the installation history updating mechanism has been greatly improved. This will improve the cleanliness of future removals of system software, among other things. The recovery mechanism is no longer necessary. Normally, repeating a failed operation is sufficient to correct any problems there may be in the history. Attempts to remove non-empty directories are retried with each new installation or removal until the directory is successfully removed. (SCR 7255, 7415, 7440, 7525) o The restrictions on installing subsystems under IRIX have been greatly reduced. (SCR 7586, 8241) o Disk space requirements are now based on a more complete perspective, and are relatively accurate even when the online installation history is incorrect. (SCR 7828) o The disk space accounting for configuration files is now handled in a more accurate manner. (SCR 7934) o The miniroot now borrows /usr/etc/resolv.conf, along with /etc/sys_id and /etc/hosts, when configuring the network. (SCR 8244) o The installation tool is now more flexible about file system configurations, and will never make a new file system without prompting for confirmation. (SCR 8527) 6.4 Bug_Fixes_to_Networking o Entering a on a line with other characters preceding it now no longer causes a logout. (SCR 2960) o A ptc opened with O_NDELAY no longer hangs the program when data is written to the pty. (SCR 6033) o The tty line discipline now behaves like those in other STREAMS implementations. A writer on a slave pty is blocked when it overruns the reader on on the matching controlling pty. Previous releases discarded the excess characters. (SCR 6034) o /etc/netgroups can now contain comments. (SCR 6687) o ftp core no longer dumps core when the open command is not given an argument. (SCR 7017) o tftpd now is more secure. It is configured by default to restrict access to files in certain directories. All options are documented in the tftpd(1M) manual entry. (SCR 7115, 9176, 9046) o ruptime(1C) now reports the load average statistics for the local machine. (SCR 7155) o The tty driver line discipline was changed to give better response to the interrupt character. (SCR 7238) o When using telnet to access a /bin/sh account on an IRIS running 4D1-3.2, typing the interrupt character would end the telnet session. This has been fixed. (SCR 7319) o The CMC ENP-10 Ethernet driver now reports late collisions. (SCR 7510) o When an IP address cannot be converted to a name, rusers(1C) now prints the address in standard dot notation. (SCR 7615) ___________________________________________/No man is a failure who has friends Trevor Paquette ICBM:51'03"N/114'05"W|I accept the challange, body and soul, {ubc-cs,utai,alberta}!calgary!paquette|to seek the knowledge of the ones of old paquette@cpsc.ucalgary.ca | - engraved on the Kersa Blade of Esalon   Received: from vmb.brl.mil by VMB.BRL.MIL id aa02098; 25 Jun 90 23:51 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac01973; 25 Jun 90 23:41 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa01958; 25 Jun 90 23:28 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa11529; 25 Jun 90 22:43 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA18954; Mon, 25 Jun 90 19:17:36 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 01:11:45 GMT From: Trevor Paquette Organization: The University of Calgary Subject: Irix 3.3 Bug_Fixes Part 2/2 Message-Id: <1990Jun26.011145.24886@calgary.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL o The uucp(1C), uucico(1M), and cu(1C) manual entries now document the read permission on /dev/ttyd* (required for cu and uucp to run.) (SCR 7829) o The kernel IP layer now uses arpwhohas() to guard against duplicate IP address claims. (SCR 8130) o In a previous release, uucico would sometimes dump core. This was caused by a divide by zero, and has been fixed. (SCR 8611) o The select(2) manual entry has been changed to accurately reflect selectable devices. o The poll(2) manual entry has been changed to include non-STREAMS devices. (SCR 8042) o When shutting down a system, last(1) no longer lists users still logged in. (SCR 8639) o The 4D1-3.3 version of finger(1C) now includes the same functionality as the BSD version. (SCR 8750) o The ndbm(3B) manual entry now correctly states the allowed size of entries (1024 bytes). (SCR 8811) o A bug in the DUART drivers that could cause the final block of data to be partially dropped has been fixed. (SCR 8933) o Internet daemons telnetd, rlogind, and ftpd now print /etc/issue before prompting for a password. (SCR 8936) 6.5 Bug_Fixes_to_the_Distributed_Graphics_Library o With the improvements to the IRIX implementation of the fork(2) system call, the DGL will now work with substantially larger applications. (SCR 7771) 6.6 Bug_Fixes_to_Mail_Programs o /bin/mail (ATT mail) used to check for a lock file when called with the -e flag. Since the purpose of the lock file is to prevent corruption of the maildrop file by insuring that only one mail process can write to the mail file at a time, and since /bin/mail called with the -e flag simply checks for the existence of mail in the maildrop file and returns without writing to it, any attempt to create a lock file in this circumstance is wrong. Not only is this behavior logically incorrect, it is potentially harmful since the existance of a legitimate (or even a bogus) lock file could cause f2/bin/mail -e to hang for several minutes. This behavior has been removed, and /bin/mail -e no longer checks for lock files. (SCR 2708) o /bin/mail previously limited the name of the mail lock file to 13 characters. Since the name of the lock file is derived by appending .lock to the user's login name, this behavior created an effective (and unnecessary) limit of 8 characters for login names which could be used with the /bin/mail program. The limit has been increased to 255 characters. (SCR 5727) o /bin/mail and /usr/sbin/Mail previously ignored the existance of Reply-To: header fields in messages. RFC 822 requires that the contents of any such Reply-To: header field be used as the return address when replying to mail. Both mail programs have been modified to comply with the RFC 822 requirement. (SCR 5993) o /usr/sbin/Mail suffered from poor parsing of message list arguments to Mail commands (such as print and delete). Parsing of message list arguments has been corrected. Please see ``Specifying messages'' in the mail_bsd(1) manual entry for a complete description of supported message list argument formats. (SCR 7409, 7953) o /usr/sbin/Mail previously used relative pathnames when dealing with mail files other than the standard maildrop. Since the Mail program supports the cd command while running, the use of relative pathnames by Mail was an unwise practice which occasionally led to unfortunate results such as deleting the wrong mailfile. /usr/sbin/Mail has been modified to resolve all user- supplied pathnames to be absolute pathnames. This practice ensures that Mail will correctly identify the files it is dealing with. (SCR 7949) o /usr/sbin/Mail uses From lines in the mail file to determine the boundaries between messages. The standard format for such From lines is ``From sender-address date''. Messages containing From lines with null sender address fields but properly formed date fields are occasionally present on the Internet (and probably other places as well). The presence of such a message in a user's mail file would confuse the Mail program and cause it to consider the entire text of the offending message to be part of the preceding message in the mail file. /usr/sbin/Mail has been modified to consider From date to be a legitimate variation on the From line and will correctly detect the boundries of such messages. (SCR 8137) o /usr/sbin/Mail had numerous problems parsing RFC 822- style mail addresses. /usr/sbin/Mail has been modified to correctly parse any RFC 822- compliant addresses. (SCR 8162, SCR 7456) o /usr/sbin/Mail did not treat comment lines (lines beginning with #) in the file $HOME/.mailrc correctly. Mail has been modified to ignore comment lines. (SCR 8196) o A security hole in sendmail that allowed remote users to write to world-writable files has been fixed. (SCR 9157) o /usr/sbin/Mail now expands aliases in the To: field prior to determining whether to use the local signature (.lsignature) or the remote signature (.rsignature). (SCR 6843) o /usr/sbin/Mail now handles From lines with additional DST fields, such as MET. (SCR 7950) 6.7 Bug_Fixes_to_Demos o The revolution demo is no longer shipped. (SCR 5827, 5837, 6535) o There was a visual problem in the light demo involving the border of the moving scene when rotating in MTV mode. (SCR 4135) o Manual entries for the dog(6D) and shadow(6D) demos were previously not included. They are now included. (SCR 4383, 6491) o Several manual entries for demos that weren't shipped with the system were being installed. These manual entries are no longer installed. (SCR 7763, 7952) o There were several typographical and spelling errors in the information screens for the radiosity demo. These have been corrected. (SCR 8099) o An error message previously appeared on the console when running the Buttonfly Show Image program. (SCR 8438) o There was a bug in the old -T option of the osview program. osview no longer supports this option. (SCR 6187) o An extraneous file (/usr/demos/.workspace/database) was being installed in the /usr/demos directory. This has been corrected. (SCR 7451) 6.8 Bug_Fixes_to_4Dgifts o The most recent version of kermit is now shipped as part of 4Dgifts. It allows you to transfer files over serial lines. (SCR 3581) o Manual entries have been added for the following programs under /usr/sbin: gamcal, hist, icut, imged, imgexp, iset, istat, izoom, mousewarp, rle, scrsave, showci, verbatim}f1. (SCR 5830) o /usr/people/4Dgifts/examples/Fortran/Makefile was zero length, causing a make executed at the top level fail. This has been corrected. (SCR 5851) o The 4Dgifts and .4sight files were owned by bin. Now they are correctly owned by guest (SCR 5990) o ipaste did not work for image files which had a colormap type of CM_DITHERED. The images were displayed as though they were black and white. This has been corrected. On 4D series machines, ipaste was written to operate on RGB (24-bit) images or SCREEN (16-bit starting at zero) images but not on the old 3000 machine version of DITHERED (8-bit starting at some offset) images. See the NOTE regarding fromdi at the bottom of the ipaste(1G) manual page. (SCR 7433, 6060) o The goem1 type declaration in 4Dgifts/examples/grafix/curve3.c has been corrected. (SCR 7701) o imged is now installed with the eoe2.sw.gltools subsystem. (SCR 7712) o The dslib manual entry incorrectly stated that the files dstab.c and dslib.c are located in /usr/people/4Dgifts/examples/devices/inquire.c. The manual entry has been corrected to reflect their correct location in /usr/people/4Dgifts/examples/devices. (SCR 8068) o Errors in the way libimage handled calls to the iopen routine have been corrected. An error handling function was added to the libimage library whose default behavior is the same as the current, but which can be changed for programs that know how to deal with the various errors. There is now a routine called i_seterror that allows you to install your own libimage error handler. (SCR 8117) o The overlay program in 4Dgifts could leave graphics drawn in the overlay bitplanes if its window was moved or killed. This has been fixed. (SCR 9038) 6.9 Bug_Fixes_to_Graphics o The graphics subsystem was killed when too many v3f calls existed between bgnpoint and endpoint. (SCR 5431) o getcpos has always returned the character position in absolute screen coordinates. However, its arguments used to be typed as Screencoord * implying a window- relative value. They are now short *. o greset would try to load the first 8 color map entries of the color map for the current drawing mode. It now calls drawmode(NORMALDRAW) first. o The setmonitor(3G) manual entry incorrectly listed the HZ30_SG option as HZ30_SG. This has been corrected. (SCR 3122) o Incorrect FORTRAN references in the manual entries of the advanced lighting routines have been corrected. (SCR 3403) o In a previous release, the drawmode manual entries incorrectly specified the argument to be logical. These manual entries now correctly specify the argument to be integer. (SCR 4282) o On GT/GTX systems, blink(3G) with rate of -1 did not turn off blinking as was documented in the manual entry. Specifying a blink rate of -1 now turns blinking off. (SCR 4350) o The defpup(3G) manual entry now lists the %n format option. (SCR 4353) o Problems using unqdevice(INPUTCHANGE) have been corrected. (SCR 5548) o The defrasterfont manual entry has been corrected to say that the ``offset'' field in the Fontchar structure is an index into the raster array, not a byte offset. (SCR 6040) o Multiple calls to blink would crash the graphics pipeline. This has been fixed in the microcode. (SCR 6139) o getmcolor used to work only in color index mode. It was fixed to work in RGB mode as well. (SCR 6621) o Context switching sometimes caused screen mask clipping of characters to fail. This has been fixed. (SCR 7346) o Negative arguments to rectzoom used to cause the graphics to crash. This has been fixed. (SCR 7348) o When picking returns a large amount of data, it could overflow and trash the name stack. This has been fixed. (SCR 7783) o Previously, backface polygon removal used an algorithm that did not work correctly in all cases. This has been fixed. (SCR 7831, 9237) o rectread in double buffer mode on an 8-bit Personal IRIS used to return the wrong values. This has been fixed. (SCR 8468) o In some cases, light source 0 would get unbound after drawing a scene. This has been fixed. (SCR 8697) o On a Personal IRIS with an RE1 raster engine (GR1.1 in the hardware inventory), a large rectwrite with a rectzoom factor greater than one used to fail. This has been fixed. (SCR 8699) o On the Personal IRIS, endfeedback used to return -1 if the feedback buffer overflowed. It now behaves as the other IRIS-4D models and returns the size of the buffer. (SCR 8709) o Depthcued, flat shaded polygons previously displayed in a single color. Now the color is recomputed for each polygon. (SCR 8838) o On a Personal IRIS with an RE1 raster engine (GR1.1 in the hardware inventory), outlined polygons failed to reset the line stipple for the first line segment. This has been fixed. (SCR 8890) o Backfacing polygons used to generate hits during picking. This has been fixed. (SCR 8910) o After an endpoint command, vertex commands used to continue to draw points. This has been fixed. (SCR 8954) o On a Personal IRIS with an RE1 raster engine (GR1.1 in the hardware inventory), rectwrites in combination with screen mask clipping used to fail. This has been fixed. (SCR 8890) o On a Personal IRIS with an RE1 raster engine (GR1.1 in the hardware inventory), stippled lines with a width greater than one did not draw correctly. This has been fixed. (SCR 9073) o rectread followed by rectwrite used to fail if the Z buffer was the readsource. This has been fixed. (SCR 9178) o On a Personal IRIS with an RE2 raster engine (GR1.2 in the hardware inventory), outlined polygons drawn with stippled lines with a width greater than one did not draw correctly. This has been fixed. (SCR 9228) o On a Personal IRIS with Turbo graphics, clipping of large polygons (more than 40 vertices), would sometimes cause the graphics to crash. This has been fixed. (SCR 9253) o On a Personal IRIS with Turbo graphics, drawing a polygon with no vertices would cause the graphics to crash. This has been fixed. 6.10 Bug_Fixes_to_4Sight o The menus in 4Sight were too sensitive. Small mouse motions would make the pull right menus disappear. (SCR 4035) o The ownerships of the files in /usr/lib/fmfonts were changed from root/bin or root/sys to bin/bin. (SCR 4106) o The psview(1) manual entry was corrected to match the functionality of the psview program. (SCR 4275, SCR 6048, SCR 6300, SCR 7289) o A range check error test was added for the third parameter of the getinterval operator, to prevent server failure when that number is negative. (SCR 4428) o The say program's unix process did not terminate when its window was killed. This was eliminated when some reference counting errors were fixed. Not reproduceable in 3.1D. (SCR 4434) o The default /etc/hosts file would attempt to route messages for the local hostname through the ethernet driver even when the driver wasn't lbooted in the kernel. This would result in the 4Sight toolchests not appearing. (SCR 4510) o A PostScript program with more gsaves than grestores could crash 4Sight. (SCR 4819) o Some graphics state operators, such as setrgbcolor would generate a typecheck error when given real numbers with more than 3 decimal places. (SCR 4873) o Using very large (> 300 pixels high) characters could crash the NeWS server. (SCR 4942) o Function invocation using defpup("title %t %F|...) did not work properly. (SCR 4982) o If the token operator was given an executable array, the NeWS server crashed with a bus error. (SCR 5108) o Drawing very small arcs using the arc or arcn operators crashed 4Sight. (SCR 5227, SCR 5228) o Certain arcs drawn with the arcn operator were drawn incorrectly and crashed 4Sight. (SCR 5229) o Drawing a dashed line from somewhere on the sceeen to a very large positive x coordinate crashed 4Sight. (SCR 5248) o A program using fprintf with a large number of arguments crashed the 4Sight when it was run as a part of a long series of PostScript tests. (SCR 5300) o Dragging whole windows rather than outlines caused problems on certain systems and could crash 4Sight. The crash has been fixed and 'Flip Drag' has been removed from the WindowChest because it doesn't work very well on certain 4D models. (SCR 5638) o The thick rubberbanding style lines would break into 2 separate lines when the window was being resized. The thick lines weren't used at all when moving a window. (SCR 5744) o The maximum number of windows that can be opened simultaneously was changed from 50 to 255 for the Personal Iris. (SCR 5920) o 4Sight hung after running psview. (SCR 5962) o Modifying a file while it was being displayed by psview hung the NeWS server. psview has been changed to terminate when the file it is displaying is changed. (SCR 6001) o The consize function was unimplemented in 3.0 and 3.1. It is implemented and working in the environment manager release. (SCR 6006) o While resizing a window, windows beneath the cursor could get input focus. This has been corrected. (SCR 6019) o Running the NeWS horse race program from comp.window.news crashed the 4Sight when it was run 15-20 times. (SCR 6081) o When a window's upper left corner was off the screen and the window was stowed, it's icon could be completely off the screen. (SCR 6188) o System performance slowed considerably after using psview intensively. (SCR 6413) o Attempting to open more than the maximum number of windows now results in a detectable failure. (SCR 6782) o Running a program that mallocs all the available memory can crash the 4Sight. (SCR 6967) o When in overlay text mode, wsh did not clear the overlay planes when stowed to an icon. (SCR 7213) o 4Sight sometimes hung or terminated abnormally when psview was used to display PostScript files imported from external sources. (SCR 7241) o Lines were not properly clipped in an imagebackground program on a Personal Iris. (SCR 7242) o The psview process did not exit when the window process terminated. (SCR 7311) o The default colormap is now documented in the 4Sight Programmer's Guide. (SCR 7337) 6.11 Bug_Fixes_to_the_WorkSpace o When a file is dragged to a new window where it overwrites an existing file, the new icon now lands where you put it (instead of at the bottom of the directory view). (SCR 7564) o When dirview is used to open a directory that does not exist, the user is informed via a notifier box, rather than a message to the console window. (SCR 7627) o WorkSpace now updates when a dangling symbolic link changes to point to a real file. (SCR 7695) o WorkSpace list views now update when the date on a file changes (if they are sorted by date). (SCR 7758) o There is a shell script in /usr/people/4Dgifts/examples/WorkSpace called shellDevice, which can be placed in /etc/transferDevice. You can then run the transferManager, select it as an active device, and access from the WorkSpace popup transfer rollover the ability to open a shell window on any selected directory. (SCR 8217, SCR 8394) o If directory views are stowed when the WorkSpace is shut down, they will be stowed when it starts up again. (SCR 8254) o Double clicking on the folder of a stowed directory view now unstows the directory view. (SCR 8264) o Files mounted from NFS now show up even if they are not exported with root privileges, or with completely open read priveleges. (SCR 8759) o When viewing as list and sorting by date, WorkSpace now resorts its list if the date on a file is changed (even if the file is only touched). o If you request to remove a file and WorkSpace can't put it in the designated dumpster directory, WorkSpace asks you if you want the file removed completely. Previously, this also turned off your preference for using the dumpster, so that anything you removed from then on was permanently removed instead of being sent to the dumpster. o On an altdrag twice to the same directory, the icon formerly disappeared although the file actually existed. o Sort by name on WorkSpace now sorts by local name, not full path name. o WorkSpaces umask is now set correctly, so that files created by WorkSpace mirror the users umask. They used to be set to completely open permissions. o You can now select more than 50 files and drop them on an executable. o If the typer cannot type an item correctly, it now identifies it as unknown, instead of as a random type. o There were several performance improvements in moving files around between directories. o If you own a directory and its contents, you can drag it or remove it even if you do not have write permission on its contents. (This supports the same behavior as the shell.) 6.12 Bug_Fixes_to_the_X_Window_System o The Xsgi man page has been redone. Further, we use traditional X man pages from the MIT tree. All X man pages and installation have been reworked. (SCR 5040, 7848) o Xsgi now reads fonts in a traditional X manner. It can read X fonts as well as fontmanager fonts. Several fontmanager interaction and lsfont interaction bugs have been fixed. (SCR 5549, 8598, 8599) o Xsgi now performs pixel-perfect rendering in all known cases. This involves a necessary performance loss in certain arenas. (SCR 7307, 7312, 7867, 8706) o X performs an ld -r on all internally built libraries, shrinking their size considerably. There is no functional change. (SCR 7851) o X header files have not changed and are consistent with the standard MIT distribution. (SCR 8224) o Xsgi startup and server defaults are configurable from the command line or from NeWS startup via xstart(1). Note that xstart(1) is intended for use from NeWS, not from the command line. (SCR 8370) o X sends correct keypress events for Alt keys. (SCR 8433) o Xsgi uses SystemV utmp(4), which fixes the bug. (SCR 8868) o Xsgi default visual type is PseudoColor. Xsgi exports two visuals: PseudoColor and StaticColor. Which is the default is configurable from the command line. (See SCR 8370, above.) Programmers who create colormaps matched to (nondefault) visuals should ensure that their windows and colormaps have the same visual type. (SCR 8944) o XPutImage has been repaired in the server. Also, primitives which descend to SetSpans to a window now correctly render in all known cases. SetSpans is completely hidden from the programmer. (SCR 9275) ___________________________________________/No man is a failure who has friends Trevor Paquette ICBM:51'03"N/114'05"W|I accept the challange, body and soul, {ubc-cs,utai,alberta}!calgary!paquette|to seek the knowledge of the ones of old paquette@cpsc.ucalgary.ca | - engraved on the Kersa Blade of Esalon   Received: from vmb.brl.mil by VMB.BRL.MIL id ab02571; 26 Jun 90 0:59 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa02468; 26 Jun 90 0:42 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa02411; 26 Jun 90 0:28 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa10639; 25 Jun 90 18:13 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA00798; Mon, 25 Jun 90 14:50:57 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 25 Jun 90 21:48:34 GMT From: Jason Heirtzler Organization: Boston University Subject: booting SGIs single-user Message-Id: <59466@bu.edu.bu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL We're about to place some PIs in a "publically" accessible area and we need to keep random people from booting the machines into single-user mode (without knowing the root password.) What sort of tricks can you do? It's IRIX 3.3 if it matters. ------------------------------------------------------------------- Jason Heirtzler (617) 353-2780 jdh@bu-pub.bu.edu Information Technology Boston University ..!bu.edu!bu-pub!jdh   Received: from vmb.brl.mil by VMB.BRL.MIL id ac02571; 26 Jun 90 0:59 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ab02468; 26 Jun 90 0:42 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab02411; 26 Jun 90 0:28 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa10641; 25 Jun 90 18:13 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA00665; Mon, 25 Jun 90 14:49:39 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 25 Jun 90 21:18:18 GMT From: Keith Goldfarb Organization: Rhythm & Hues, Inc., Hollywood Subject: Replacing the perspective matrix. Message-Id: <696@celia.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL The perspective() call will load a matrix onto the transformation stack if in single matrix mode or replace the current projection matrix if in viewing mode. The loadmatrix() call will load a matrix onto the transformation stack no matter what the mode. What I want to do is to replace the current projection matrix (while in viewing mode) with one of my choosing with a call like loadmatrix(). Does such a call exist? K. -- Keith Goldfarb Rhythm & Hues Claudia is Everywhere. celia!keith@usc.edu celia!keith@tis.llnl.gov ...mlogic!celia!keith I got nothing. Too bad. But I'm happy 'cause that's all I have.   Received: from vmb.brl.mil by VMB.BRL.MIL id ad02571; 26 Jun 90 0:59 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ac02468; 26 Jun 90 0:42 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa02432; 26 Jun 90 0:29 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa11312; 25 Jun 90 21:43 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA15949; Mon, 25 Jun 90 18:31:57 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 01:20:49 GMT From: Al Dobryden Organization: University of Michigan EECS Dept., Ann Arbor Subject: A/D board Message-Id: <2803@zipeecs.umich.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Does anybody have experience installing an A/D converter board in a 4D/210? I'm particularly interested in anybody that has interfaced a Data Translation board. Please reply by email, if you can help or if you have a similar problem. My address is ald@crim.eecs.umich.edu. thanks, Al The University of Michigan Artificial Intelligence Laboratory   Received: from vmb.brl.mil by VMB.BRL.MIL id aa02726; 26 Jun 90 1:10 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa02571; 26 Jun 90 0:59 EDT Received: from adm.brl.mil by VMB.BRL.MIL id aa02472; 26 Jun 90 0:33 EDT Received: from TROUT.NOSC.MIL by ADM.BRL.MIL id aa17294; 26 Jun 90 0:28 EDT Received: from ucsd.edu by trout.nosc.mil (5.59/1.27) id AA19319; Mon, 25 Jun 90 17:18:49 PDT Received: from chema.ucsd.edu by ucsd.edu; id AA25783 sendmail 5.64/UCSD-2.1-sun via SMTP Mon, 25 Jun 90 17:18:45 -0700 for @nosc.mil:info-iris@brl.mil Received: by chem.chem.ucsd.edu (5.51) id AA14835; Mon, 25 Jun 90 17:17:48 PDT Date: Mon, 25 Jun 90 17:17:48 PDT From: Steve Dempsey Message-Id: <9006260017.AA14835@chem.chem.ucsd.edu> To: info-iris@BRL.MIL Subject: 3.3 for the Personal IRIS Does anybody know when IRIX 3.3 will be released for the Personal IRIS? We just received a 340VGX with 3.3 installed and there are some nice features I like to use on my PI. -------------------------------------------------------------------------------- Steve Dempsey (619) 534-0208 Dept. of Chemistry Computer Facility, B-014 INTERNET: sdempsey@ucsd.edu University of Calif. at San Diego BITNET: sdempsey@ucsd La Jolla, CA 92093 UUCP: ucsd!sdempsey   Received: from vmb.brl.mil by VMB.BRL.MIL id ac00788; 26 Jun 90 15:39 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ad00479; 26 Jun 90 15:28 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ac00225; 26 Jun 90 15:10 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa02749; 26 Jun 90 14:48 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA17449; Tue, 26 Jun 90 11:19:54 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 25 Jun 90 22:15:47 GMT From: sgi!shinobu!odin!tristram@ucbvax.berkeley.edu Organization: Subject: Re: Panel Library from NASA Ames Message-Id: References: <9006240725.16686@godzilla> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9006240725.16686@godzilla> mg@GODZILLA.CGL.RMIT.OZ.AU (Mike Gigante) writes: As I read this, I could feel the words changing before my eyes... here are my new ones and some reactions. I see someone from SGI suggested the panel library for a customer's application in a previous message in this group. reasonable suggestion, the Library works, and is cheap and fast. I have used the panel library extensively - having written nearly a dozen applications using it. I am very grateful to the author of the panel library (Dave Tristram) - it has been a valuable prototyping tool. thank you... However, I don't really think it is worth starting from scratch with it at this point. SGI have previously announced in this group that a SGI toolkit will be available as part of the next (post 3.3) release. This toolkit may be available as part of some future (post 3.3) release. This toolkit ^^^ ^^^^ ^^^^^^ the SGI toolkit has not been announced, and can not be counted on in any particular release, although it stands to reason SGI will be providing some kind of enhanced programming environment in the future. will have C++, C and f77 bindings. Pascal I'm not sure about. From the little I know of this toolkit, it seems like it should be much better than the panel library and the resultant applications will fit nicely into the SGI environment. very likely The panel library suffers from not haveing any documentation and also being there is a hundred and fifty page user's guide, and way too many demos... very clumsy for many things. To make things work as you would like is please, somewhat clumsy for a few things... :-) difficult, some common interaction styles almost impossible. The code ends up being quite ugly as your application will need to muck with the internals of the panel library a *lot* if you wish to have fine control. kind of depends on one's ideas of internals, doesn't it? Much mucking can be performed without ever touching the event code (panel.c). (Admittedly, the PL has a farily restrictive idea of what people are going to be doing with their mouses, I'm sure there are a lot of implementations of modal dialog boxes out there...) That is given the high overheads to start with thge panel library and that The Panel Library was designed with minimum programmer interface as a top priority, it's possible to add a button to a gl program in a few minutes. I feel the mental overhead, while detectable, is extremely low. Startup performance overhead is limited only by how fast the window manager can get a window on the screen. a significantly better toolkit is not more than 6 months away, I can't recommend it. Care to make a wager? However, until now it has been our only real choice given the SGI's X implementation and no alternative interaction stuff. I have to agree that there are few truly professional GUI toolkits for the Iris, see Iris Universe marketing rag for a commercial sampling. (Sorry, I guess that the NeWS stuff has been there for a while, but I've never tried it. Wise move since it will dissappear eventually..) Needling from me aside, your main points are well taken. The Panel Library is a lightweight, easy to use, but not quite first-class UI toolkit. X-based toolkits are harder to use, and give you, well, X. SGI's entry will, at the very least, be well-documented, robust, professionally designed code, and provide direct access to all the power of the hardware. Mike Gigante, RMIT Advanced Computer Graphics Centre David Tristram tristram@sgi.com --   Received: from vmb.brl.mil by VMB.BRL.MIL id ab01120; 26 Jun 90 15:57 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ad00788; 26 Jun 90 15:45 EDT Received: from adm.brl.mil by VMB.BRL.MIL id aa00739; 26 Jun 90 15:36 EDT Received: from ucbvax.Berkeley.EDU by ADM.BRL.MIL id aa08276; 26 Jun 90 14:16 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA16807; Tue, 26 Jun 90 11:11:02 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 01:17:12 GMT From: Gary Tarolli Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Replacing the perspective matrix. Message-Id: <62846@sgi.sgi.com> References: <696@celia.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <696@celia.UUCP>, keith@celia.UUCP (Keith Goldfarb) writes: > The perspective() call will load a matrix onto the transformation > stack if in single matrix mode or replace the current projection > matrix if in viewing mode. The loadmatrix() call will load a > matrix onto the transformation stack no matter what the mode. > > What I want to do is to replace the current projection matrix > (while in viewing mode) with one of my choosing with a call > like loadmatrix(). Does such a call exist? > The way to do this is to get into MPROJECTION mode, using mmode(MPROJECTION); and then use loadmatrix, and return to the previous mmode. In MVIEWING mode, loadmatrix will load the model-view matrix, so that is why you must get into MPROJECTION mode. This should work, but there may be a bug on the GT/GTX machines (I think this was reported and fixed in the just released 3.3 version). Gary Tarolli   Received: from vmb.brl.mil by VMB.BRL.MIL id aa01307; 26 Jun 90 16:07 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ae00788; 26 Jun 90 15:46 EDT Received: from adm.brl.mil by VMB.BRL.MIL id ab00739; 26 Jun 90 15:36 EDT Received: from ucbvax.Berkeley.EDU by ADM.BRL.MIL id aa08311; 26 Jun 90 14:18 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA16357; Tue, 26 Jun 90 11:05:05 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 25 Jun 90 22:08:50 GMT From: dave "who can do? ratmandu!" ratcliffe Organization: Silicon Graphics Inc., Entry Systems Division Subject: Re: WorkSpace error: "can't connect with file access monitor" Message-Id: <9442@odin.corp.sgi.com> References: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article , broderic@u2.rutgers.edu (Alfred J. Broderick) writes: |> |> I am unable to get the Workspace program to start up. The error |> message that I get is... |> |> Can't connect with File Access Monitor (fam) |> |> I have tried to start /usr/etc/fam by hand with no luck. Could |> someone tell me what I am doing wrong? Also, could someone describe |> what the fam is? |> |> I have a 4D70G running 3.2.1 the file access monitor (fam) is a service that keeps WorkSpace and the file system in contact. it takes updated info about/in the file system (including whatever updating (creating/removing files/dirs/etc) is occurring) and delivers it to WorkSpace. it is not tied to clock ticks except over nfs mounts. below is an internal posting i put together from different SGI engineer's responses to this problem when 3.2 first came out: ------------------------------------------------------------------------------- cust was finding one of 2 problems occuring when, on their 3.2 4D machine, they tried to either fire up ypbind AFTER workspace was already running, OR trying to invoke workspace AFTER ypbind was already running. 1. if workspace is already running & ypbind is invoked, the system hangs solid 2. if ypbind is already running and they try to invoke workspace they get the error "can't connect with file access monitor". The problem was that their YP server machine was not a 3.2 IRIX machine (in this case it wasn't even an IRIS--it was a SUN). below are three ultimate knowledge pieces regarding this situation: ------------------------------------------------------------ Fact: If a site is running 3.2 software, its YP master machine must also be running the 3.2 version of /etc/rpc, which has the services needed for file system monitoring contained within it. ------------------------------------------------------------ Subject: Re: 3.2 MR ws & yp prob On the YP master, after making sure /etc/rpc is the 3.2 version (with sgi_fam and sgi_toolkitbus entries), they should cd /usr/etc/yp make rpc and watch things happen. ----------------------------------------------------------- Subject: Re: 3.2 MR ws & yp prob ALL this yp and workspace stuff : Two problems have been around : 1 - If the yp server is not running 3.2, the /etc/rpc database for everyone who is using yp will not have sgi_fam and sgi_toolkitbusd in it. To fix this, the following lines should be added to /etc/rpc on the server: sgi_toolkitbus 391001 sgi_fam 391002 Then, you should ( on the server ) cd /usr/etc/yp; make rpc 2 - Some people on larger networks have had trouble with autoworkspace. Basically, if the network is large and slow enough, the yp server is not resolved by the time the system tries to start the workspace, and so the workspace fails to start ( since it can't connect to fam, since yp won't tell it how to....) A second or so later, if you try to start the workspace from the system menu, it starts up fine (since yp has by then gotten it's act together...) -- daveus rattus yer friendly neighborhood ratman KOYAANISQATSI ko.yan.nis.qatsi (from the Hopi Language) n. 1. crazy life. 2. life in turmoil. 3. life out of balance. 4. life disintegrating. 5. a state of life that calls for another way of living.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa02938; 26 Jun 90 17:52 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa00479; 26 Jun 90 15:27 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa00125; 26 Jun 90 15:09 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa02508; 26 Jun 90 14:19 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA16134; Tue, 26 Jun 90 11:02:03 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 25 Jun 90 20:53:40 GMT From: Reuel Nash Organization: Silicon Graphics, Inc., Mountain View, CA Subject: night program mods Message-Id: <62821@sgi.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I have modified the night program to fix a few obscure bugs. See comments below. ---- cut here --- (and signature) /* Original program from: howardl@landis.csd.sgi.com (Howard Look) Here's yet another little background program. It looks like the sky at sunset. It will only run properly on 24 or more bit plane systems. Enjoy. night enhancements uunet!alberta!calgary!paquette Jun-19-1990: Trevor Paquette - hacked up data structures for 'square stars' - added twinkle feature Please email any enhancements to me. Jun-25-1990: Reuel Nash - changed timers from timeout/signal handler to queued device TIMER1: you can't do graphics reliably from inside a signal handler and main, too. Change overall timer rate with noise() call. - removed unecessary stuff from trand(). - the pixel readback scheme to find out what color to "blink" to won't work if there are windows obscuring the background on startup. This is now calculated by the program from background colors and Y1 and Y2. See get_sky(). - stars don't need to be shaded. Use single packed color for each star and background of star. - added "comets". (really more like shooting stars). Compile it with: cc -o night night.c -lm -lgl_s -lc_s -lmalloc */ #include #include #include #include #include /* a star is actually a "square" p1(x,y,r,g,b,br,bg,bb) p4(r,g,b,br,bg,bb) +----------------------------+ |<---^------ rad ----------->| | | | | rad | | | | | V | +----------------------------+ p2(r,g,b,br,bg,bb) p3(r,g,b,br,bg,bb) */ typedef struct _dummy { float x, y, rad; /* location and radius of star */ long color; /* color of star */ long back; /* background color behind star */ } starstruct; #define Y1 0.0 #define Y2 0.2 #define SMALL_MAX_BLINK 128 #define BIG_MAX_BLINK 5 int numgroups = 10; /* num of groups of 256 stars */ int numbig = 200; /* num of big stars */ starstruct *small_stars = NULL; /* pointer to locations of small stars */ starstruct *big_stars = NULL; /* pointer to locations of big stars */ starstruct *small_blinker[SMALL_MAX_BLINK]; /* at most MAX_BLINK stars blink at a time */ starstruct *big_blinker[BIG_MAX_BLINK]; /* at most MAX_BLINK stars blink at a time */ /* colors now global so get_sky() works */ int orange[] = {255,72,0}; int blueish[] = {0,110,189}; int black[] = {0,0,0}; int red[] = {255,0,0}; main(argc, argv) int argc; char *argv[]; { int gid, skyevent(), i; short val; long dev; char *malloc(); for(i = 0; i < SMALL_MAX_BLINK; i++) { small_blinker[i] = NULL; } for(i = 0; i < BIG_MAX_BLINK; i++) { big_blinker[i] = NULL; } small_stars = (starstruct *) malloc(numgroups*256*sizeof(starstruct)); big_stars = (starstruct *) malloc(numbig*sizeof(starstruct)); #ifndef DEBUG imakebackground() ; #endif /* foreground();*/ gid = winopen("") ; RGBmode(); gconfig(); shademodel(GOURAUD); ortho2(0.0, 1.0, 0.0, 1.0) ; qdevice(TIMER1); noise(TIMER1, 30); qenter(REDRAW,gid); while (1) { dev = qread(&val) ; switch(dev) { case REDRAW: draw_background(); break; case TIMER1: skyevent(); break; } } } draw_background() { int i,j, skyevent(), col[3]; static int first = 1; float v1[2],v2[2],v3[2],v4[2]; register int counter; v1[0] = v4[0] = 0.0; v2[0] = v3[0] = 1.0; v1[1] = v2[1] = 0.0; v3[1] = v4[1] = Y1; bgnpolygon(); c3i(red); v2f(v1); v2f(v2); c3i(orange); v2f(v3); v2f(v4); endpolygon(); v1[1] = v2[1] = Y2; bgnpolygon(); c3i(orange); v2f(v4); v2f(v3); c3i(blueish); v2f(v2); v2f(v1); endpolygon(); v3[1] = v4[1] = 1.0; bgnpolygon(); c3i(blueish); v2f(v1); v2f(v2); c3i(black); v2f(v3); v2f(v4); endpolygon(); if(first) { setup_sky(); first = 0; } counter = 0; for(j=0; jcolor); v2f(&small_blinker[i]->x); endpoint(); } } if(big_numstars) { for(i = 0; i < big_numstars; i++) { bgnpolygon(); v1[0] = big_blinker[i]->x; v1[1] = big_blinker[i]->y; v2[0] = big_blinker[i]->x; v2[1] = big_blinker[i]->y - big_blinker[i]->rad; v3[0] = big_blinker[i]->x + big_blinker[i]->rad; v3[1] = big_blinker[i]->y - big_blinker[i]->rad; v4[0] = big_blinker[i]->x + big_blinker[i]->rad; v4[1] = big_blinker[i]->y; cpack(big_blinker[i]->color); v2f(v1); v2f(v2); v2f(v3); v2f(v4); endpolygon(); } } if(comet.active) { bgnline(); cpack(get_sky(comet.y0)); v2f(&comet.x0); cpack(get_sky(comet.y1)); v2f(&comet.x1); endline(); comet.active = 0; } small_numstars = trand(1, SMALL_MAX_BLINK); /* blink between 1 and 256 stars */ for(i = 0; i < small_numstars; i++) { starnum = trand(0, numgroups*256); small_blinker[i] = &small_stars[starnum]; bgnpoint(); cpack(small_stars[starnum].back); v2f(&small_stars[starnum].x); endpoint(); } big_numstars = trand(1, BIG_MAX_BLINK); /* blink between 1 and 256 stars */ for(i = 0; i < big_numstars; i++) { starnum = trand(0, numbig); big_blinker[i] = &big_stars[starnum]; bgnpolygon(); v1[0] = big_blinker[i]->x; v1[1] = big_blinker[i]->y; v2[0] = big_blinker[i]->x; v2[1] = big_blinker[i]->y - big_blinker[i]->rad; v3[0] = big_blinker[i]->x + big_blinker[i]->rad; v3[1] = big_blinker[i]->y - big_blinker[i]->rad; v4[0] = big_blinker[i]->x + big_blinker[i]->rad; v4[1] = big_blinker[i]->y; cpack(big_blinker[i]->back); v2f(v1); v2f(v2); v2f(v3); v2f(v4); endpolygon(); } if((event >= 1) && (event <= 10)) { /* shooting star event */ comet.x0 = ((float)trand(0,32767))/32767.0; comet.y0 = ((float)trand(0,32767))/32767.0; comet.x1 = comet.x0 + ((float)trand(0,32767))/327670.0; comet.y1 = comet.y0 + ((float)trand(0,32767))/327670.0; bgnline(); cpack(0x808080); v2f(&comet.x0); cpack(get_sky(comet.y1)); v2f(&comet.x1); endline(); comet.active = 1; } if((event >= 2) && (event <= 10)) { /* satilite event (see above)*/ } } setup_sky() { register int i, j; register int counter; long tmp; long trand(); long get_sky(); /* setup big stars , should be done BEFORE small stars */ for(j =0; j Subject: IRIX 3.3 installed To: info-iris@BRL.MIL X-VMS-To: X%"info-iris@brl.mil" Message-ID: <9006260306.aa00655@VGR.BRL.MIL> Hallo, this is from a man who received 3.3 and couldn't wait to install it before the current bug-list is out. First of all, the software is running and (if you don't count a deffective 'dev' tape, see below) the installation was smooth and easy. I like the new c compiler for it's prototyping, the enhanced gr_osview is great and the new gr_top would replace any need for a VAX if it only had a bar-mode for the top-CPU users. Now it comes to the problems. Up to now I've found the following glitches. Most of them are just annoying and easy to surround, but the last one is in my eyes severe: 1.) I cannot find the man pages for malloc(3x) in the printed documentation. It used to be near malloc(3m), but it disappeared. 2.) The terminfo entry for the good old vt-100 has changed. The smkx (ks) and rmkx (ke) entries, which should switch on/off the alternate keypad don't. Using infocmp, my editor and tic I changed: smkx=\E[?1h to smkx=\E[?1h\E= and rmkx=\E[?1l to rmkx=\E[?1l\E> 3.) About two weeks ago I posted a message about the 3.3 make, that I had the chance to see on a pre-3.3 installation. Now I repeat it, because in my 3.3 make shows the same behaviour. In the top-level makefile for a program system we define our CFLAGS and pass them to related lower-level makefiles using the following construct: (cd somedir; make "CFLAGS = $(CFLAGS)") That used to work under 3.2 and did not work under 3.3. Now the surprising story begins: To find a fix I tryed to find wisdom in the make man-page. There I found that I could pass the CFLAGS also via the environment variables. To test that, I entered (just for joke): setenv CFLAGS bla Suddenly (instead of passing 'bla' as the CFLAGS) our 'real' CFLAGS from the top-level makefile appeared in the lower-level makes. Btw. we are using the c-shell. 4.) This the (for me) most severe problem. Telnet does not work, if you use it from a vt100 directly connected to one of the serial ports. If you try to connect to a VAX running VMS and CMU/TPC, it fills the screen with (at least it looks like) parity errors. Trying to connect to other unix systems (including localhost) ends at the 'username:' prompt. Surprisingly, if you first rlogin (that works) to localhost, than telnet works as expected (even with that VAX). Now to the installation: After installing eoe1, eoe2 and NFS, I had to discover that the dev tape was broken after ~19MB of the tape. Trying to read the tape on three different drives proved that it was really the fault of the cartridge (aren't the tapes tested befored shipping ?). As it was friday noon and the next SGI office is in Munich (400 km from Darmstadt) there was no chance to get a replacement on the same day (here I have to thank the people in Munich: they managed to send me the tape on monday morning, despite the fact that it was practically to late to distcp the dev-tape and ship it on the same day). So what should I do? Seeing myself as a bit more experienced than a novice user, I decided not to install the full old 3.2 software, but install the 3.2 dev-kit over the 3.3 eoe1+2. That worked quite well while inside 'inst'. But when I tried to compile c-programs, I had to discover that something was wrong with the include files. After a closer look at the directory I saw that the /usr/include/sys tree was from the 3.3 release. Next thing I've done was to install the complete 3.2 include tree and that worked. It seems that during the installation of the eoes' some parts of the include files are installed, which are not part of the 'dev' kit. My suggestion for future releases: Is it possible to include the full include tree in the 'dev' kit? That would mean some overhead on the tape, but would ease situations like mine. Regards Martin Knoblauch TH-Darmstadt Physical Chemistry 1 Petersenstrasse 20 D-6100 Darmstadt, FRG BITNET:   Received: from vmb.brl.mil by VMB.BRL.MIL id aa03805; 26 Jun 90 4:36 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa03697; 26 Jun 90 4:26 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa03638; 26 Jun 90 4:16 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa00785; 26 Jun 90 4:05 EDT Received: from MCCLB0.MED.NYU.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 5374; Tue, 26 Jun 90 03:55:23 EDT Received: from mcirps2.med.nyu.edu by MCCLB0.MED.NYU.EDU; Tue, 26 Jun 90 03:55 EDT Received: by mcirps2.med.nyu.edu (5.52/890607.SGI) (for @mcclb0.med.nyu.edu:info-iris@brl.mil) id AA07184; Tue, 26 Jun 90 04:27:34 DSD Date: Tue, 26 Jun 90 04:27:34 DSD From: root%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: Do somthing with a directory besides viewing To: info-iris@BRL.MIL Message-id: <9006261127.AA07184@mcirps2.med.nyu.edu> X-Envelope-to: info-iris@brl.mil If I click on a directory below, I want to startup my appliation, not open the directory. Workspace does not seem to recognize these two matches. What am I doing bad ? TYPE SkullDirectory MATCH dircontains(".SkullDir"); # MATCH glob("D.Skulls"); CMD OPEN echo $LEADER ICON { include("SkullDir.fti"); } -- +-----------------------------------------------------------------------------+ | karron@nyu.edu Dan Karron | | . . . . . . . . . . . . . . New York University Medical Center | | 560 First Avenue \ \ Pager <1> (212) 397 9330 | | New York, New York 10016 \**\ <2> 10896 <3> | | (212) 340 5210 \**\__________________________________________ | +-----------------------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id aa05209; 26 Jun 90 7:27 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa05129; 26 Jun 90 7:17 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa05095; 26 Jun 90 7:09 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa01153; 26 Jun 90 7:01 EDT Received: from DDATHD21.BITNET by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 6612; Tue, 26 Jun 90 07:00:21 EDT Received: from BR2.THD.DA.D.EUROPE by DDATHD21.BITNET via GNET with RJE ; 26 Jun 90 13:01:05 Date: Tue, 26 Jun 90 12:58:23 +0200 (Central European Summer Time) From: Knobi der Rechnerschrat Subject: Inquiring window-status To: info-iris@BRL.MIL X-VMS-To: X%"info-iris@brl.mil" Message-ID: <9006260701.aa01153@VGR.BRL.MIL> Hallo, is there a way for an arbitrary process to determine whether any window in the system is stowed or not? Is there a way to unstow windows from outside (there must be one, because it's in the window-menues). Any help is appreciated. Regards Martin Knoblauch TH-Darmstadt Physical Chemistry 1 Petersenstrasse 20 D-6100 Darmstadt, FRG BITNET:   Received: from vmb.brl.mil by VMB.BRL.MIL id aa05908; 26 Jun 90 8:04 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa05534; 26 Jun 90 7:43 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa05447; 26 Jun 90 7:33 EDT Received: from prandtl.nas.nasa.gov by VGR.BRL.MIL id aa01200; 26 Jun 90 7:24 EDT Received: Tue, 26 Jun 90 04:23:26 -0700 from csduts1.lerc.nasa.gov by prandtl.nas.nasa.gov (5.61/1.2) Received: Tue, 26 Jun 90 07:25:32 EDT by csduts1.lerc.nasa.gov (5.51/LeRC(1.0)) Received: Tue, 26 Jun 90 07:47:28 EDT by avelon.lerc.nasa.gov (5.52/LeRC(1.0)) Date: Tue, 26 Jun 90 07:47:28 EDT From: Tony Facca Message-Id: <9006261147.AA14316@avelon.lerc.nasa.gov> To: info-iris%brl.mil@prandtl.nas.nasa.gov Subject: Re: 3.3 for the Personal IRIS > Does anybody know when IRIX 3.3 will be released for the Personal IRIS? > We just received a 340VGX with 3.3 installed and there are some nice > features I like to use on my PI. Yes, I'd like to know this as well. Also, would some one from SGI please answer the following: -- What rev of the OS is currently shipping for the 4D/25TG's? -- What rev of the OS is currently shipping for the 4D/3x0VGX's? Thanks. -- ----------------------------------------------------------------------------- Tony Facca | fsfacca@avelon.lerc.nasa.gov | phone: 216-433-8318 ----------------------------------------------------------------------------- You are at Witt's end. Passages lead off in *all* directions.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa10500; 26 Jun 90 11:56 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa10182; 26 Jun 90 11:45 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa10088; 26 Jun 90 11:35 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa02003; 26 Jun 90 11:29 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA05233; Tue, 26 Jun 90 08:29:00 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 13:16:36 GMT From: mcsun!cernvax!chx400!urz.unibas.ch!doelz@uunet.uu.net Organization: University of Basel, Switzerland Subject: 3.3. shipment date wanted Message-Id: <1990Jun26.141636.804@urz.unibas.ch> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In Switzerland, the promised 3.3 bulk shipment date is announced as 'late September'. I'd appreciate hins on SGI ISO distribution policy and where 3.3 is shipped already. Reinhard   Received: from vmb.brl.mil by VMB.BRL.MIL id aa12486; 26 Jun 90 13:45 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa11842; 26 Jun 90 13:24 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa11713; 26 Jun 90 13:13 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa02247; 26 Jun 90 12:59 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA11083; Tue, 26 Jun 90 09:51:28 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 07:26:20 GMT From: Greg Chien Organization: Illinois Institute of Technology, Chicago Subject: Turn-key application under 4Sight Message-Id: <3895@iitmax.IIT.EDU> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I'm trying to create a turnkey demo account that should run only one application -- loads a few slides onto the frame buffer and then logs out automatically when done. I've been trying to put the program in the /etc/passwd entry and/or modifying the user.ps file to spawn a unix process without success. Can anyone tell me the proper way to allow a user to run just one application under 4sight? Is the command "echo exitcleanly | psh" appropriate for auto logout with no side-effect or is there any better solution? Thanks in advance Greg Chien Manager, Design Processes Laboratory Institute of Design Illinois Institute of Technology idchien@iitmax.iit.edu Phone: (312) 567-6461 Fax : (312) 567-3779   Received: from vmb.brl.mil by VMB.BRL.MIL id aa00788; 26 Jun 90 15:38 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab00479; 26 Jun 90 15:27 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa00225; 26 Jun 90 15:10 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa02726; 26 Jun 90 14:48 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA18279; Tue, 26 Jun 90 11:33:18 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 18:10:22 GMT From: Betsy Zeller Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Outpaced in WorkSpace Message-Id: <62891@sgi.sgi.com> References: <9006230547.AA26618@mcirps2.med.nyu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9006230547.AA26618@mcirps2.med.nyu.edu> root@MCIRPS2.MED.NYU.EDU writes: > Can I make icons with quickpaint ? > Can I paint (how) patterns, help icons, and other things I would want ? No, not on this release. The iconsmith (mail request to monica@sgi.com) is the only tool we have that will save your favourite picture as the kind of geometry that WorkSpace can use. > What about an include file functionality for the *.ftr files. Check out the documentation (README files) that comes with the iconsmith. This will explain how to utilize the include functionality. > >I got rid of root and the file tree below it.I only want it to come up with >my application root, and the FIRST LEVEL DIRECTORIES directly below it. >NOTHING MORE under that. The user would pick the directory. The application >would pick the files in the directory. > When you make the template file, you should lay it out just the way that you want the user to see it. It is important to realize that the blue WorkSpace window reflects both the template file ( or the users own view have created from that), and the contents of the user's ~/WorkSpace directory. Remember, if you are using the template to prune the user view of files, you need to remove the user's file will be created the next time that the user runs, but its contents will now be restricted to what was specified in the template file, as well as what the user has since done to customize the view. > I want to give my application a unique feel The customization facilities available will let you specialize how your icons look, how they behave when they are opened, and what menu choices are available from them. > How can I run multiple workspace environment together ? At > different times ? At the same time ? There is currently one WorkSpace process available per user. If you want to run another WorkSpace process as another user, login or su - as that user, and start WorkSpace from that shell. > I would like to have private prototype environment that would not reach > out of my local area (like to /usr/lib/filetype and /usr/lib/workspace). If you want to do this, look at the file /usr/lib/filetype/Makefile.personal. This will tell you about how to set up your own customizations in $HOME/.workspace/local $HOME/.workspace/install $HOME/.workspace/system $HOME/.workspace/default You can add in new type rules and behaviours here, and they will override the system ones. Any type or behaviour you don't specify will come from the system ftr files. > I would also like to have multiple environments to develop, one for > each major application. Many of us are feeling the need for different blue WorkSpace windows, each of which can be set up for a different task. Look for it in a future release. > I would like lots of things, only I need to figure out what I would > need/want. > Eventually, your workspace would engulf my applications in some seamless > fashon. This is certainly what we are working towards - its great to hear that we share the same vision. Keep those questions and requests coming. Betsy Zeller betsy@sgi.com   Received: from vmb.brl.mil by VMB.BRL.MIL id ab00788; 26 Jun 90 15:39 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac00479; 26 Jun 90 15:28 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab00225; 26 Jun 90 15:10 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa02741; 26 Jun 90 14:48 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA17292; Tue, 26 Jun 90 11:18:19 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 04:54:00 GMT From: Rob Warnock Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Irix 3.3 Additions Message-Id: <62861@sgi.sgi.com> References: <1990Jun26.010821.24567@calgary.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL One addition really made me *really* happy was: 4.1.3 New_Commands_and_Utilities o Keyboard auto-repeat rate may be adjusted on workstations using the new keywarp(6) command. Try "keywarp 20 2", and watch your cursor keys (and "hjkl" in "vi") fly! -Rob ----- Rob Warnock, MS-9U/510 rpw3@sgi.com rpw3@pei.com Silicon Graphics, Inc. (415)335-1673 Protocol Engines, Inc. 2011 N. Shoreline Blvd. Mountain View, CA 94039-7311   Received: from vmb.brl.mil by VMB.BRL.MIL id ac02605; 26 Jun 90 17:31 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa02382; 26 Jun 90 17:20 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa02365; 26 Jun 90 17:06 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa03644; 26 Jun 90 16:44 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA26903; Tue, 26 Jun 90 13:35:41 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 20:22:01 GMT From: James Zurlo Organization: Mechanical Engineering, Carnegie Mellon, Pittsburgh, PA Subject: Sun raster to rgb? Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I've been seeing all sorts of converters from other graphic formats to the Silicon Graphics rgb format, but I haven't seen a Sun raster to rgb converter. Does it exist and if so where? Thanks in advance. Jim jz0t+@andrew.cmu.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa03497; 26 Jun 90 19:52 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa03449; 26 Jun 90 19:41 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa03439; 26 Jun 90 19:35 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa03219; 26 Jun 90 15:14 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA20465; Tue, 26 Jun 90 12:02:24 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 18:47:23 GMT From: usc!zaphod.mps.ohio-state.edu!uwm.edu!dogie.macc.wisc.edu!vms.macc.wisc.edu!bruggink@ucsd.edu Subject: newuser questions Message-Id: <3901@dogie.macc.wisc.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL 1) From what machine can I FTP the utah raster toolkit? Are there other pd routines/utilities available for the silicon graphics? 2) Does a utility exist to dump the screen to a postscript file? --dennis bruggink bruggink@vms.macc.wisc.edu bruggink@ccc.nersc.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa03831; 26 Jun 90 21:30 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa03625; 26 Jun 90 20:27 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa03620; 26 Jun 90 20:19 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa04217; 26 Jun 90 20:00 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA11129; Tue, 26 Jun 90 16:57:54 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 23:48:48 GMT From: "Alfred J. Broderick" Organization: Rutgers Univ., New Brunswick, N.J. Subject: 4Dgifts and 4sight server from 3.3 tapes. Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Hi! A few weeks ago, Mark Callow mentioned that there were some incompatibilities between 4Sight 3.3 and the ~4Dgifts/.4sight/*.ps files. One of the problems was that the frame title gets obscured by a black line. Unfortunately I did not save Mark's message when it passed through the newsgroup. Was there a fix for this problem? All of my users have used the *.ps files in the 4Dgifts directory as the basis for their 4Sight environments. Thank you. Alfred ps. I have asked three questions in this newsgroup and each time I received prompt and correct information from both SGI employees and other SGI enthusiasts. I really appreciate you all helping me out. Keep up the good work! ~ ~ ~   Received: from vmb.brl.mil by VMB.BRL.MIL id aa04773; 27 Jun 90 1:13 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab04688; 27 Jun 90 1:02 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab04655; 27 Jun 90 0:49 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa04890; 27 Jun 90 0:30 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA28213; Tue, 26 Jun 90 21:15:28 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 20:31:17 GMT From: Thant Tessman Organization: Silicon Graphics Inc. Subject: Re: Replacing the perspective matrix. Message-Id: <9514@odin.corp.sgi.com> References: <696@celia.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <696@celia.UUCP>, keith@celia.UUCP (Keith Goldfarb) writes: > The perspective() call will load a matrix onto the transformation > stack if in single matrix mode or replace the current projection > matrix if in viewing mode. The loadmatrix() call will load a > matrix onto the transformation stack no matter what the mode. > > What I want to do is to replace the current projection matrix > (while in viewing mode) with one of my choosing with a call > like loadmatrix(). Does such a call exist? I don't think switching between MPROJECTION and MVIEWING changes the values in the matrices. (Only going into MSINGLE does that.) So I think you can just say: mmode(MPROJECTION); loadmatrix(m); mmode(MVIEWING); Not having actually done this myself, I don't guarantee the validity of this information, but I think its right. thant "Ask not for whom the booth tolls..."   Received: from vmb.brl.mil by VMB.BRL.MIL id ab04773; 27 Jun 90 1:13 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac04688; 27 Jun 90 1:02 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ac04655; 27 Jun 90 0:49 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa04893; 27 Jun 90 0:31 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA28581; Tue, 26 Jun 90 21:21:34 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 01:04:35 GMT From: Bean Anderson Organization: Silicon Graphics Inc. Subject: Re: G0 libraries Message-Id: <9555@odin.corp.sgi.com> References: <9006240711.16666@godzilla>, <3513@umbc3.UMBC.EDU> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <3513@umbc3.UMBC.EDU>, bernie@umbc5.umbc.edu (Bernard J. Duffy) writes: > > ... > > Does this mean programs performing dynamic loading will bomb if they > weren't compiled -G 0 or would they just yield the wrong numbers/ fault > results? If they bomb, I'll just leave them off the system until some > user complains about their program not working due to a dynamic loading > error (hopefully an informative sys-err message on failure). I currently > don't have any flavors of lisp running, but might have KCL in the future > (added to our machines by our KCL support person). > > Programs performing dynamic loading will not necessarily "bomb" if they weren't compiled -G 0. The main thing is that the dynamically loaded modules were compiled -G 0; if not, the dynamically loaded module will end up overlaying data space in the GP region. Bean   Received: from vmb.brl.mil by VMB.BRL.MIL id ac04773; 27 Jun 90 1:13 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ae04688; 27 Jun 90 1:02 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa04658; 27 Jun 90 0:49 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa04906; 27 Jun 90 0:32 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA28497; Tue, 26 Jun 90 21:20:25 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 00:18:45 GMT From: Paul Haeberli Organization: Silicon Graphics, Inc., Mountain View, CA Subject: convert sun images to IRIS image file format Message-Id: <62933@sgi.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL The following program converts from all known sun rasterfile images to IRIS image file format. If you have a sun file that doesn't work, pls let me kno. paul haeberli paul@sgi.com begin 777 fromsun M`6``!R:'^H\````````````X`````\"$,$!4'KX2`=`P0%&@``"`ACX2`=(^%@'B/AH!P#!``?``` M```,$!2D`$`@(0````T#X``(`````">]``@#X``(`"#X(0/@``@````````` M`">]_U"OM``P*($``P"@H"&OO@`XK[\`/*^U`#2OL@`HK[,`+!0@``2OL0`D M/!,0`!````DF%@!`40``(KZ(`J#P%$`".A@`$)*4`)`_@!%0"8"`A#!`4I"0$ M``&/IP"H)Z0`9"0%``$/X`1:)`8`((^F`'"/N`!TCZ\`;#P.6:8USFJ5C[X` M:(^R`'B/L0"`)`$`"*^N`&2OI@"0K[@`F`,`J"$0P0`-KZ\`H"0!`!@0P0`* M`````"0!``$0P0`'`````#P%$``DI0`\#^`$5`)@("$,$!2D)`0``18@``L` M````C[D`D"0!``@7(0`'`````#P%$``DI0!<#^`$5`)@("$,$!2D)`0``1H@ M`$T`````)`$``P(A`!JOL`!,/`00`8^G`*@DA)YP)`8``0``@!(/X`1:`@`H M(3P$$`&/IP"H)(2?<`(`*"$/X`1:)`8``3P$$`&/IP"H)(2@<`(`*"$/X`1: M)`8``8^P`$P0```T`````(^E`)`,$`+B`\`@(8^H`*```````0(`&0``2!(5 M```"```````'``TD`?__%0$`!#P!@``5(0`"```````&``VOJ0!T`2@`&@`` M4!(0```IKZH`F(^K`'2/K`"@``````%L`!H5@``"```````'``TD`?__%8$` M!#P!@``580`"```````&``T``&@2$```&:^M`)B/I0"0#!`"X@/`("$0```4 MKZ(`F#P%$``DI0!\`F`@(0_@!%0"0#`A#!`4I"0$``$0```,CZX`D!)`_\V/ MI0"0)`$``1)!_]^/JP!T)`$``A)!_^R/I0"0$`#_[@````"/K@"0)`$``17! M``T`````CH0`"(^O`*`D&``!K[@`&*^^`!`GA8`4)`8!`20'``(,$`N]KZ\` M%!````P`0*`ACH0`"(^Y`*`D"0`#KZD`&*^^`!`GA8`8)`8!`20'``,,$`N] MK[D`%`!`H"$D`0`"%D$`@X^H`*"/J`"8CZH`H*^P`$P!"@`9``"`$@_@!((" M`"`ACZ0`J`!`D"$"H"@A`$`P(0P0`G@"`#@ACZL`H`)`B"$98`!N``"8(:^V M`$BOMP!$/!<0`#P6$`"/H@"0/!(0`#P5$``FM5YP)E(^<";6?G`0``!1)O<. M<`(@("$"0"@A#!`&70/`,"&/K0"@CZP`F`&S,",DQO__`H`@(0)`*"$``#@A M#!`2G`(LB"&/H@"0$```2X^I`*`"("`A`D`H(0*@,"$"P#@A#!`"0J^^`!"/ MKP"@CZX`F`'S@",F$/__`@`P(0*`("$"0"@A```X(0P0$IP"+H@A`H`@(0*@ M*"$"`#`A#!`2G"0'``$"@"`A`L`H(0(`,"$,$!*<)`<``H^B`)`0```NCZD` MH``>*("/IP"H`+XH(P(@("$/X`1:)`8``8^X`)@"X"`A`D`H(0*@,"$"P#@A MK[X`$`P0`F(".(@AC[D`H`*`("$#,X`C)A#__P(`,"$"0"@A#!`2G```."$" M@"`A`J`H(0(`,"$,$!*<)`<``0*`("$"P"@A`@`P(0P0$IPD!P`"CZ(`D!`` M``N/J0"@)`$``1!!_Z\"("`A)`$`"!!!_[P"("`A)`$`&!!!_]4`````CZD` MH"9S``$":0@J%"#_]"0!``&/M@!(C[<`1`````"/L`!,$```=0````"/J`"@ M``"8(1D``'$`````K[8`2*^W`$0\%Q``/!80`#P2$``\%1``)K5><"92/G`F MUGYP)O<.*("/IP"H`+XH(P+@ M("$/X`1:)`8``0+@("$"0"@A`J`P(0+`."$,$`)BK[X`$(^L`*`"@"`A`9.` M(R80__\"`#`A`D`H(0P0$IP``#@A`H`@(0*@*"$"`#`A#!`2G"0'``$"@"`A M`L`H(0(`,"$,$!*<)`<``A````R/K@"@CZT`D"0!``$1H?^L`````"0!``@1 MH?^[`````"0!`!@1H?_7`````(^N`*`F]`+"/HP`0``````!@$"$00``:)&/__SP($`$\"1`! M/`H0`25*H'`E*9]P)0B><)".````8!`A`0YX(9'X```D8___I+@``)"9```D MI0`"`3E8(9%L```DQ@`"I,S__I"-```DA``!`4UP(9'/```DYP`"%$#_[J3O M__X#X``(`````(^C`!```````&`0(1!``!`D8___D(X``"2$``&DK@``D(\` M`"2$``&DSP``D)@```!@$"$D8___)(0``22E``(DQ@`").<``A1`__*D^/_^ M`^``"``````GO?^@K[``$`#`&"$`X$`A`*"`(0$0<".OLP`8KZ0`8(^G`&`` M;I@A`&AX(:^_`"2OM```-"OL@`XDF,``":4__\4=0`H)G,``1Z```8``````L`@(0_@!%0" MX"@A#!`4I"0$``&2<```)I3__Q(``!HF]_^BOOP`4`*`8(0_@!!```RA`C[\` M%">]`!@#X``(`````">]_^BOOP`4`,`8(0_@!`P``S!`C[\`%">]`!@#X``( M`````">]_^@4H``%K[\`%`P0`NP`P"@A$```&8^_`!0HP0`(%"``#@#`$"$D MQO_X*,$`"*2%``"DA0`"I(4`!*2%``:DA0`(I(4`"J2%``RDA0`.$"#_]22$ M`!``P!`A$$``!B3&__\`P!`A),;__Z2%```40/_\)(0``H^_`!0GO0`8`^`` M"``````HP0`(%"``;P#`$"&$@@````````1!``0H00$`$```"*2@```H00$` M%"``!`!`&"$0```")`,`_P!`&"&DHP``A((``@`````$00`$*$$!`!````BD MH``"*$$!`!0@``0`0!@A$````B0#`/\`0!@AI*,``H2"``0`````!$$`!"A! M`0`0```(I*``!"A!`0`4(``$`$`8(1````(D`P#_`$`8(:2C``2$@@`&```` M``1!``0H00$`$```"*2@``8H00$`%"``!`!`&"$0```")`,`_P!`&"&DHP`& MA((`"``````$00`$*$$!`!````BDH``(*$$!`!0@``0`0!@A$````B0#`/\` M0!@AI*,`"(2"``H`````!$$`!"A!`0`0```(I*``"BA!`0`4(``$`$`8(1`` M``(D`P#_`$`8(:2C``J$@@`,``````1!``0H00$`$```"*2@``PH00$`%"`` M!`!`&"$0```")`,`_P!`&"&DHP`,A((`#@`````$00`$*$$!`!````BDH``. M*$$!`!0@``0`0!@A$````B0#`/\`0!@AI*,`#B3&__@HP0`()(0`$!`@_Y0D MI0`0`,`0(1!``!,DQO__A((````````$00`$*$$!`!````BDH```*$$!`!0@ M``0`0!@A$````B0#`/\`0!@AI*,```#`$"$DQO__)(0``A1`_^\DI0`"`^`` M"``````GO?_H)`$``13!``6OOP`4#!`%\0#@,"$0``!6C[\`%"0!__\4P0`& M*.$`"`P0!B<`X#`A$```3X^_`!0HX0`(%"``/@#@$"&$K@``A)@```#.`!F$ MB@`"A(X`!"3G__@HX0`()(0`$"2E`!```'@2`?C((:29__"$J/_R``````#( M`!F$B/_V``!($@$J6"&DB__RA*S_]```````S``9A(S_^```:!(!KG@AI(__ M](2X__8``````-@`&828__H``,@2`RA((:2)__:$JO_X``````#*`!F$BO_\ M``!8$@%L:"&DC?_XA*[_^@``````S@`9A([__@``>!(!^,@AI)G_^H2H__P` M`````,@`&0``2!(!*E@AI(O__(2L__X``````,P`&0``:!(!KG@A$"#_Q:2/ M__X`X!`A$$``#"3G__^$N```A(@```#8`!D`X!`A).?__R2E``(DA``"``#( M$@,H2"$40/_VI(G__H^_`!0GO0`8`^``"``````4A0"0*.$`""CA``@4(`!W M`.`0(82N``"$N``"`<8`&H2H``2$J@`&A*P`"!3```(```````<`#20!__\4 MP0`$/`&``!7!``(```````8`#82N``H4P``"```````'``TD`?__%,$`!#P! M@``7`0`"```````&``T4P``"```````'``TD`?__%,$`!#P!@``5`0`"```` M```&``T4P``"```````'``TD`?__%,$`!#P!@``500`"```````&``T4P``" M```````'``TD`?__%,$`!#P!@``5@0`"```````&``T4P``"```````'``TD M`?__%,$`!#P!@``5P0`"```````&``TDY__X)*4`$```>!*DK__P``````,& M`!J$N/_\`````!3```(```````<`#20!__\4P0`$/`&``!!*DK__\A)C__@`````#!@`:%,```@`` M````!P`-)`'__Q3!``0\`8``%P$``@``````!@`-*.$`"```R!(0(/^-I+G_ M_@#@$"$00``3).?__X2(````X!`A`08`&A3```(```````<`#20!__\4P0`$ M/`&``!4!``(```````8`#23G__\DA``")*4``@``2!(40/_OI*G__@/@``@` M````CZ,`$``````H80`(%"``J0!@$"&4N```E(X``)3)````&,B``SC((0`. M>(``&YX(0`/>0`#.,@C``E0@``9R(`![G@C`4E0(0`*4(`!^4`A`4E0 M(0$*6"$`"V("I.P``)2X``*4C0`"`!AX@)3)``(!^'@A``UP@``/>,`!S7`A M``YQ``'X>",`"4"```]X@`'-<",!"4`A``A`@`'/R"$!"4`A`RA0(0`*6@*D MZP`"E+@`!)2,``0`&'"`E,D`!`'8<"$`#&B```YPP`&L:"$`#6D``=AP(P`) MR(``#G"``:QH(P,IR"$`&EX(0`/>(`!C7`A`>EX(0'/R"$`&4("I.@`!I2X``B4B@`(`!A@@)3)``@! MF&`A``I8@``,8,`!:E@A``M9``&88",`"7"```Q@@`%J6",!R7`A``YP@`%L M:"$!R7`A`:YX(0`/R@*D^0`(E+@`"I2(``H`&%B`E,D`"@%X6"$`"%"```M8 MP`%(4"$`"E$``7A8(P`):(``"UB``4A0(P&I:"$`#6B``4M@(0&I:"$!C7`A M``YZ`J3O``J4N``,E)D`#``84("4R0`,`5A0(0`90(``"E#``1E`(0`(00`! M6%`C``E@@``*4(`!&4`C`8E@(0`,8(`!"E@A`8E@(0%L:"$`#7("I.X`#)2X M``Z4CP`.E,D`#@`80(`!&$`A``_(@``(0,`#+\@A``E8@``9R0`!&$`C`6E8 M(0`+6(``"$"``R_((P,H4"$!:5@A`4M@(21C__@H80`(``QJ`J3M``XDA``0 M)*4`$"3&`!`0(/]:).<`$`!@$"$00``<)&/__Y2X``"4C@``E,D````8R(`# M.,@A``YX@``9R,`![G@A``E0@``/>0`#.,@C`4E0(0`*4(``&YX(P'Y M0"$!25`A`0I8(0`+8@(`8!`A)&/__Z3L```DA``")*4``B3&``(40/_F).<` M`@/@``@`````*,$`"!0@`"<`P!`AA*\``(2.``"$F0`"`<_`(:28``"$J``" MA(H`!`,H2"&DB0`"A*L`!(2-``8!2V`AI(P`!(2N``:$F``(`:YX(:2/``:$ MN0`(A(D`"@,90"&DB``(A*H`"H2,``P!*E@AI(L`"H2M``R$CP`.`8UP(:2. M``R$N``.),;_^"C!``@!^,@AI)D`#B2$`!`0(/_<)*4`$`#`$"$00``*),;_ M_X2(``"$J0```,`0(0$)4"&DB@``),;__R2E``(40/_X)(0``@/@``@````` M*,$`"!0@`"<`P!`AA*\``(2.``"$F0`"`<_`(Z28``"$J``"A(H`!`,H2".D MB0`"A*L`!(2-``8!2V`CI(P`!(2N``:$F``(`:YX(Z2/``:$N0`(A(D`"@,9 M0".DB``(A*H`"H2,``P!*E@CI(L`"H2M``R$CP`.`8UP(Z2.``R$N``.),;_ M^"C!``@!^,@CI)D`#B2$`!`0(/_<)*4`$`#`$"$00``*),;__X2(``"$J0`` M`,`0(0$)4".DB@``),;__R2E``(40/_X)(0``@/@``@`````),+__P1!``(` M0`@A)"$`!P`!$,,D0@`!`(`X(1A``#@``!@A)`8`_Y#D```DYP`!,(X`@!'` M``,`````$````J2@``"DI@``,(\`0!'@``,`````$````J2@``*DI@`",)@` M(!,```,`````$````J2@``2DI@`$,)D`$!,@``,`````$````J2@``:DI@`& M,(@`"!$```,`````$````J2@``BDI@`(,(D`!!$@``,`````$````J2@``JD MI@`*,(H``A%```,`````$````J2@``RDI@`,,(L``1%@``,`````$````J2@ M``ZDI@`.)&,``0!B""H4(/_+)*4`$`/@``@`````),+__P1!``(`0`@A)"$` M!P`!$,,D0@`!`(`X(1A``#<``!@AA.X`````("$IP0"`$"```@`````D!`"` MA.\``@`````IX0"`$"```@`````TA`!`A/@`!``````K`0"`$"```@`````T MA``@A/D`!@`````K(0"`$"```@`````TA``0A.@`"``````I`0"`$"```@`` M```TA``(A.D`"@`````I(0"`$"```@`````TA``$A.H`#``````I00"`$"`` M`@`````TA``"A.L`#@`````I80"`$"```@`````TA``!)&,``0!B""HDYP`0 MH*0``!0@_\LDI0`!`^``"`````"/AX!@)[W_Z*^_`!0`@#`A%.``&P"@2"$D M!`$`KZ8`&`_@!(*OJ0`7`)!,```(`!"!`-(0``21C``$H80`(%"#_^"0.``$DI0`!**$! M`*$$```4(/_P)0@``2DA``@4(``F`2`0(9#9``"0S``!`/E0(9%+````[&@A MH,L``)&N``"0SP`"H,X``0#OP"&3&0``D,H``Z#9``(`ZE@AD6P``)#-``2@ MS``#`.UP(9'/``"0V``%H,\`!`#XR"&3*@``D,L`!J#*``4`ZV`AD8T``)#. M``>@S0`&`.YX(9'X```E*?_X*2$`""3&``@0(/_=H-C__P$@$"$40``#)2G_ M_Q````JOAX!@D-D```$@$"$`^5`AD4L``"4I__\DQ@`!%$#_^:#+__^OAX!@ MC[\`%">]`!@#X``(```````%<$`$H0`"`*`((20A``$``2A#`*`P(0".&"$D M8__^)*7__Q#```H`@!`AA$0``(1O````H#`AI$\``"2E__\D0@`")&/__A3` M__BD9``"`^``"``````HP0`($"``!3P#``$\`P`!$```+S1C`0$\`P`!-&,! M`92.```DQO_X`<,`&2C!``@DI0`@)(0`$```>!*LK__@E)C_\@`````#`P`9 M``#($JRY_^24B/_T``````$#`!D``$@2K*G_Z)2*__8``````4,`&0``6!*L MJ__LE(S_^``````!@P`9``!H$JRM__"4CO_Z``````'#`!D``'@2K*__])28 M__P``````P,`&0``R!*LN?_XE(C__@`````!`P`9``!($A`@_]6LJ?_\`,`0 M(1!```HDQO__E(H```#`$"$!0P`9),;__R2$``(DI0`$``!8$A1`__BLJ__\ M`^``"`````"/HP`0`````"AA``@4(`!)`&`0(92N``"4F```E,@````.>@`! M^,@E``A,``,I4"6LZ@``E,\``I2-``*4JP`"``_$```+8@`!C7`E`=A`):SH M``24S``$E(H`!)2Y``0`#&P``!E*``$J6"4!;7@EK.\`")2N``:4R0`&E(@` M!@`.P@``"50``PC()0,J8"6L[``,E*L`")38``B4CP`(``MJ```81``!KW`E M`@`!^,@E``E4``,J6"4`#&X``6UP):Q.``"4KP`"E.T``I3*``*4B0`"``_" M```*9``#"<@E`RQ8)0`-=@`!;G@EK$\`!)2X``24[@`$E,P`!)2*``0`&$H` M``QL``$JR"4#+5@E``Y^``%OP"6L6``(E*D`!I3-``:4C``&``E2`)3O``8` M#70``4S()0,N6"4`#\8``7A():Q)``R4J@`(E,X`")2-``@`"F(`E/@`"``. M?``!C<@E`R]8)0`83@`!:5`EK$H`$)2L``J4SP`*E(X`"@`,:@"4Z0`*``_$ M``&NR"4#.%@E``E6``%J8"6L3``4E*T`#)38``R4CP`,``UR`)3J``P`&$P` M`<_()0,I6"4`"F8``6QH):Q-`!B4K@`.E,D`#I28``Z4[``.``YZ```)5``! M^,@E`RI8)21C__@`#&X``6UP)2AA``BL3@`<)$(`("2$`!`DI0`0),8`$!`@ M_Z$DYP`0`&!`(1$``!0D8___E*\``)2)``"4R@``E.T````/P@`#"<@E``ID M``,L6"4`#78``6YX)0!@0"$D8___K$\``"3G``(DQ@`")*4``B2$``(5`/_N M)$(`!`/@``@`````CZ,`$``````H80`(%"``4P!@0"&,@@``)&/_^#!.`/^D MK@````)Z`C'X`/^DV`````+,`C,I`/^DZ0``C((`!"AA``@P2@#_I*H``@`" M6@(Q;`#_I,P``@`";`(QK@#_I.X``HR"``@DA``@,$\`_Z2O``0``L(",QD` M_Z39``0``DP",2H`_Z3J``2,@O_L)*4`$#!+`/^DJ__V``)B`C&-`/^DS0`& M``)T`C'/`/^D[P`&C(+_\"3&`!`P6`#_I+C_^``"R@(S*0#_I,G_^``"5`(Q M2P#_I.L`"(R"__0DYP`0,$P`_Z2L__H``FH",:X`_Z3.__H``GP",?@`_Z3X M__J,@O_X`````#!9`/^DN?_\``)*`C$J`/^DRO_\``)<`C%L`/^D[/_\C(+_ M_``````P30#_I*W__@`"<@(QSP#_``+$`C,9`/^DS__^$"#_L*3Y__X`8$`A M$0``$21C__^,@@```&!`(3!)`/^DJ0````)2`C%+`/\``F0",8T`_Z3+``"D M[0``)&/__R2$``0DI0`"),8``A4`__$DYP`"`^``"`````"/HP`4`````"AA M``@0(``%CZ@`$(^H`!`0``!M`&!((8^H`!``````C((``"1C__@P3@#_I*X` M```">@(Q^`#_I-@````"S`(S*@#_I.H````"7@(Q;`#_I0P``(R"``0H80`( M,$T`_Z2M``(``G(",<\`_Z3/``(``L0",QD`_Z3Y``(``E8",4L`_Z4+``*, M@@`()(0`(#!,`/^DK``$``)J`C&N`/^DS@`$``)\`C'X`/^D^``$``+.`C,J M`/^E"@`$C(+_["2E`!`P2P#_I*O_]@`"8@(QC0#_I,T`!@`"=`(QSP#_I.\` M!@`"Q@(S&0#_I1D`!HR"__`DQ@`0,$H`_Z2J__@``EH",6P`_Z3,__@``FP" M,:X`_Z3N``@``GX",?@`_Z48``B,@O_T).<`$#!9`/^DN?_Z``)2`C%+`/^D MR__Z``)D`C&-`/^D[?_Z``)V`C'/`/^E#P`*C(+_^"4(`!`P6`#_I+C__``" MR@(S*@#_I,K__``"7`(Q;`#_I.S__``";@(QK@#_I0[__(R"__P`````,$\` M_Z2O__X``L(",QD`_Z39__X``E0",4L`_P`"9@(QC0#_I.O__A`@_YBE#?_^ M`&!((1$@`!4D8___C((```!@2"$P3@#_I*X````">@(Q^`#_I-@````"S`(S M*@#_``)>`C%L`/^DZ@``I0P``"1C__\DA``$)*4``B3&``(DYP`"%2#_[24( M``(#X``(`````(R#`!"4@@`&)`$`_Q!A`)L`````$&``F0`````H00`(%"`` M@0!`("&$K@``A+D``@`.>@`![G@C`>,`&@`90@`!&4`CA*H`!(2M``8`"EH` M`6I8(P`-<@`!S7`C%&```@``````!P`-)`'__Q1A``0\`8``%>$``@`````` M!@`-%&```@``````!P`-)`'__Q1A``0\`8``%0$``@``````!@`-%&```@`` M````!P`-)`'__Q1A``0\`8``%6$``@``````!@`-%&```@``````!P`-)`'_ M_Q1A``0\`8``%<$``@``````!@`-)$+_^"2E`!```,`2I+C_\(2X__@!`P`: M`!C*``,XR",48``"```````'``TD`?__%&$`!#P!@``7(0`"```````&``T` M`$@2I*G_\H2I__H!8P`:``E2``%)4",48``"```````'``TD`?__%&$`!#P! M@``500`"```````&``T``&`2I*S_](2L__P!PP`:``QJ``&L:",48``"```` M```'``TD`?__%&$`!#P!@``5H0`"```````&``T``'@2I*__]H2O__X#(P`: M``_"``,/P",48``"```````'``TD`?__%&$`!#P!@``7`0`"```````&``TH M00`(``!`$J2H__@``````4,`&@``6!*DJ__Z``````&C`!H``'`2I*[__``` M```#`P`:``#($A`@_X*DN?_^`$`@(1"``!0D0O__A*@```!`("$`"$H``2A( M(P$C`!H48``"```````'``TD`?__%&$`!#P!@``5(0`"```````&``TD0O__ M)*4``@``4!(4@/_NI*K__@/@``@`````)[W_X*^P`!@`@(`AK[\`'*^E`"2O MI@`HE@(`!H^.@"``````$<(`'@````"/CX`@`````!'@``P`````CX2`9`_@ M!(``````CX2`:`_@!(``````CX2`;`_@!(``````E@(`!@`````/X`2"``(@ M0*^"@&26!``&#^`$@@`$($"O@H!HE@0`!@_@!((`!"!`KX*`;)88``8````` MKYB`()89``H`````+R$``Q`@``@`````CZ4`)(^F`"@"`"`A#!`3O@``."$0 M```8C[\`'(^%@&2/I@`H`@`@(0P0$[X``#@ACX6`:(^F`"@"`"`A#!`3OB0' M``&/A8!LCZ8`*`(`("$,$!.^)`<``I8(``:/A(!DCX6`:(^&@&R/IP`D#!`% M):^H`!"/OP```-P````"&&`!R`````!,```0"`"`A#!`-C`(`("$"`"`A M`@`H(0P0#DDD!@"8AAD`<@`````3(``#``````P0#8P"`"`AE@@``B0!`0`Q M"?\`%2$`(@`````"`"`A#!`.A20%`@"6"@`(E@L`"@`````!2P`9``!@$@`, M:("OK0`]_^BOOP`4E(X`<``````QSP`"$>``&(^_`!2,A0"``````!"@`!2/OP`4 MC)@`?``````3```0C[\`%(2&`':$AP!X#!`2G*^D`!B/I``8`````)29``8` M````$%D`!H^_`!24B`!P`````#4)`""DB0!PC[\`%">]`!@#X``(`````">] M_^BOOP`4E(X`<``````QSP`!%>```P`````0```S)`+__XR8`(``````%P`` M!@`````,$`U,KZ0`&(^D`!@`````K((`@(2&`':4F0`(``````#9""L4(``2 M`````(2"`'B4B``*)$(``0`"%````A0#`$@(*Z2"`'@4(``'I(``=I2)`'"D M@`!X-2H`$*2*`'`0```5)`+__X2&`'8`````C(4`@(2'`'@,$!.^KZ0`&(^D M`!@`````E(L`!HR#`("$C0!V)6S__ZR#`'PEK@`!I(X`=J2,`'J48@``)&\` M`JR/`'PP0O__C[\`%">]`!@#X``(`````">]_^BOOP`4KZ4`')2.`'`````` M,<\``A7@``,`````$```/"0"__^,A0"``````!2@``@`````#!`-3*^D`!B/ MI``8`````*R"`(",A0"``````(R8`'P`````$P``(`````"$A@!VA(<`>`P0 M$IROI``8CZ0`&`````"$@@!VE)D`""1"``$``A0```(4`P!9""L4(``0I((` M=H2"`'B4B``*)$(``0`"%````A0#`$@(*Z2"`'@4(``'I(``=I2)`'"D@`!X M-2H`$*2*`'`0```0)`+__XR%`(``````E(L`!JR%`'PE;/__I(P`>H^M`!P` M````I*T``(R.`'P`````)<\``JR/`'R/H@`<`````(^_`!0GO0`8`^``"``` M```GO?_HK[\`%`"`&"$D9``8#^`$R"0&`%"/OP`4)[T`&`/@``@``````^`` M"*R%`&@GO?_8KZ4`+*^F`#``X!@ACZX`.(^O`#R/N`!`KZ0`*(^E`"B/IP`P MCZ8`+*^_`"0``"`AKZ,`$*^N`!2OKP`8#!`+YZ^X`!R/OP`D)[T`*`/@``@` M````)[W_V*^F`#``X!@ACZX`.(^O`#R/N`!`KZ4`+(^F`"R/IP`PK[\`)``` M*"&OHP`0KZX`%*^O`!@,$`OGK[@`'(^_`"0GO0`H`^``"``````GO?_8K[\` M%*^D`"@D!`"8K[``$*^E`"ROI@`P#^`$@J^G`#0`0(`A`$`@(0_@!!`D!0"8 MCZX`,`````"1SP`!`````#GX`"LO&``!$P``"*^X`"`\!!``/`40`"2E`/`/ MX`14)(0&,`P0%*0D!``!C[D`,"0!`'>3*````````!4!`%V/JP`LCZ8`+``` M```0P``4CZH`*`#`("$,$!2X)`4!MH^I`""OH@`H$2``"@`````$0``(```` M`(^D`"@,$!2P`````(^D`"P,$!3`)`4``J^B`"B/I@`L`````(^J`"@````` M!4$`"8^B`#@\!!``/`40`"2E`1@/X`14)(0&,`P0%*0D!``!CZ(`."0+`=JF M"P``CZP`-"0#``&F#``"CZT`/"Q!``*F`P`*I@,`"!0@``2F#0`&CZX`0``` M``"F#@`(+$$``Q0@``0`````CZ\`1`````"F#P`*EA@`"@`````7`P`()`D` M`Y8(``@D&0`"%0,`!:89``00```#I@,`!"0)``.F"0`$/`H`F#5*EH"N"@`, MK@``$`(`("$,$`NQ)X6`,*X``!2F``!RCZ0`*`(`*"$,$!3()`8`F"0!`)@0 M00!5CZL`(#P$$``\!1``)*4!/`_@!%0DA`8P#!`4I"0$``$0``!,CZL`((^K M`"P`````$6``#8^M`"B/K``@`````!&```0``"@A$````B0%``(``"@ACZ0` M+`P0%,``````KZ(`*(^M`"@`````!:$`!(^D`"@0``#0```0(8^D`"@"`"@A M#!`4T"0&`)@D`0"8$$$`"``````\!!``/`40`"2E`60/X`14)(0&,`P0%*0D M!``!E@8``"0!`=HPSP#_``_"```&<@(!V,@E%R$`!P`````D"``!I@@`<@P0 M#8P"`"`A$````Y8&``"F``!RE@8``"0!`=H0P0`(`````#P$$``\!1``)*4! MC`_@!%0DA`8P#!`4I"0$``&6`@`$`````"Q!``,0(``%+$$``B0)``&6`@`$ MI@D`"BQ!``(0(``$CZL`("0*``&F"@`(CZL`(``````18``%CZT`,"0,`(`0 M```,I@P`<(^M`#`D`0!RD:\````````1X0`%)!@``20.``(0```#I@X`<"08 M``&F&`!PEAD``B0!`0`S*/\`%0$`X"`*X.`(R/N``P`````),9````````%R$`&8^D M`"B6"``(E@D`"@``,"$!"0`9```H$AB@`$4````````0(0`%((`D`___C@H` MD``````!0E@AK6```(X,`)0``````8)H(21"``0`1`@J%"#_]JVC```0```V MI@``>H^D`"@D!0(`#!`4V```,"&/I``HC@4`D(^F`!P,$!30`````(^O`!P` M````$$\`"``````\!!``/`40`"2E`=`/X`14)(0&,`P0%*0D!``!A@X`<@`` M```1P``&CZ0`*(X$`)"/I0`<#!`-*X"`(0D!0(````P(0P0 M%-BN!`!L`@`0(8^_`!2/L``0`^``"">]`"@GO?_@K[\`%`"`&"&48@`&```` M```"<8(!PB`A#^`$@@`$((`40``(KZ(`'#P$$``\!1``)*4"&`_@!%0DA`8P M#!`4I"0$``&/OP`4CZ(`'`/@``@GO0`@``400P"`,"$80``,```8(93$```D M8P`!``,<```#'`,`!'("``1Z``'/P"4`8@@JI-@``!0@__8DQ@`"`^``"``` M````!1"#`(`P(1A``!0``!@A/`<`_P`#<(``SB@AC*0``"1C``$`!,(",QG_ M```$?@(`!$H``2=0)`'Y0"4``QP```,<`P$*6"4`!&8``6QH)0!B""H4(/_O MK*T```/@``@`````)[W_Z*^D`!B/I``8K[\`%`P0#6(D!0`,CZ0`&"0%``P, M$`US)(0`#(^D`!@D!0`$#!`-]`!@#X``(`````">]_^BOOP`4`*`X(0#@*"&OIP`O MI@`@CZ0`&(^F`""/IP`!1``!:DAP!VE((`!I2/ M``@`P@`9,&D`_P``(2'`'8!SP`9C(D`E```P!(#!\@A`!E`@`$H4"&-1@````````3! M``R/OP`4/`00`#P%$``DI0,()(0&,`_@!%2OI@`<#!`4I"0$``&/I@`<```` M`(^_`!0GO0`@`^``"`#`$"$GO?_HKZ8`(*^E`!ROIP`DCZ8`)(^E`""OOP`4 M#!`.)Z^D`!B/I``8CZX`))2/``B/N0`@`<\`&8R*`)",B`"4C(D`C(^G`!PD M`?__``#`$@,9*"$`!1"``4)8(0$"&"&M:0``C&8````````0P0`%`````(R, M`!0``````89H(:R-`!2L9P``C(X`C``````!QW@AK(\`C(^_`!0GO0`8`^`` M"``````GO?_H`*`8(20"``&OOP`4%&(`AP#`2"$4X@"%`````(^N`"@`@#@A M`(Y`(0"(""L0(`![`2`H(0#@("$DYP`"`.@(*Q`@`!<`````D.+__Y#O__X` M````%>(`!0````"0^````````!!8``X`````).<``0#H""L0(``*`````)#B M__^0^?_^`````!P$@*"$`X"`A).<``@#H""L0 M(``7`````)#B__^0^/_^`````!<"``4`````D/D````````060`.`````"3G M``$`Z`@K$"``"@````"0XO__D.K__@`````50O_X`````)#K````````%$O_ M]``````DY__^`.0P(Q#``#8`````*,$`?Q0@``0`!AP`$````R0#`'X`!AP` M``,<`S1L`(`H80`)I*P```##,",4(``9)*4``I"-```D8__XI*T``)".``$` M`QP`I*X``I"/``(``QP#I*\`!)"8``,H80`)I+@`!I"9``0DI0`0I+G_^)"* M``4DA``(I*K_^I"+__X`````I*O__)",__\0(/_II*S__@!@$"$D8___``,< M`!!```H``QP#`&`0(9"-```D8___``,<```#'`,DA``!)*4``A1`__BDK?_^ M%,#_S2C!`'\`X"`A).<``0#H""N0X___$"``#@#D,".0[@```&`0(17"``H` MY#`C).<``0#H""L0(``&`.0P(Y#O````````$>+_^0``````Y#`C$,``$`#H M""L`8!`A*,$`?Q0@``0`!AP`$````R0#`'X`!AP```,<`P##,".DHP``)*4` M`B2E``(4P/_TI*+__@#H""L4(/^(`.`@(22E``(`J1`C!$$``@!`""$D(0`! M``$00Q```2^DH/_^%&4`D``````4X@".`````(^X`"@`@#@A`!C(0`"90"$` MB`@K$"``@P$@*"$`X"`A).<`!`#H""L0(``7`````)3B__Z4ZO_\`````!5" M``4`````E.L````````02P`.`````"3G``(`Z`@K$"``"@````"4XO_^E.S_ M_``````5@O_X`````)3M````````%$W_]``````DY__\`.0P(P3!``(`P`@A M)"$``0`!,$,0P``V`````"C!`'\4(``$``8<`!````,D`P!^``8<```#'`,T M;@"`*&$`":"N````PS`C%"``&22E``&4CP``)&/_^*"O``"4F``"``,<`*"X M``&4F0`$``,<`Z"Y``*4B@`&*&$`":"J``.4BP`()*4`"*"K__R4C``*)(0` M$*"L__V4C?_\`````*"M__Z4CO_^$"#_Z:"N__\`8!`A)&/__P`#'``00``* M``,<`P!@$"&4CP``)&/__P`#'````QP#)(0``B2E``$40/_XH*___Q3`_\TH MP0!_`.`@(23G``(`Z`@KA./__A`@``X`Y#`CE/@```!@$"$7`@`*`.0P(R3G M``(`Z`@K$"``!@#D,".4^0```````!,B__D``````.0P(P3!``(`P`@A)"$` M`0`!,$,0P``0`.@(*P!@$"$HP0!_%"``!``&'``0```#)`,`?@`&'````QP# M`,,P(Z"C```DI0`!)*4``13`__2@HO__`.@(*Q0@_X``X"`A)*4``0"I$",0 M``">H*#__Q1E`)0`````%.4`D@````"/J@`H`(`X(0`*6$``BT`A`(@(*Q`@ M`(,!("@A`.`@(23G``0`Z`@K$"``%P````"4XO_^E.S__``````5@@`%```` M`)3M````````$$T`#@`````DYP`"`.@(*Q`@``H`````E.+__I3N__P````` M%<+_^`````"4[P```````!1/__0`````).?__`#D,",$P0`"`,`((20A``$` M`3!#$,``-@`````HP0!_%"``!``&'``0```#)`,`?@`&'````QP#-'@`@"AA M``FDN````,,P(Q0@`!DDI0`"E)D``"1C__BDN0``E(H``@`#'`"DJ@`"E(L` M!``#'`.DJP`$E(P`!BAA``FDK``&E(T`""2E`!"DK?_XE(X`"B2$`!"DKO_Z ME(___`````"DK__\E)C__A`@_^FDN/_^`&`0(21C__\``QP`$$``"@`#'`,` M8!`AE)D``"1C__\``QP```,<`R2$``(DI0`"%$#_^*2Y__X4P/_-*,$`?P#@ M("$DYP`"`.@(*X3C__X0(``.`.0P(Y3J````8!`A%4(`"@#D,",DYP`"`.@( M*Q`@``8`Y#`CE.L````````18O_Y``````#D,",$P0`"`,`((20A``$``3!# M$,``$`#H""L`8!`A*,$`?Q0@``0`!AP`$````R0#`'X`!AP```,<`P##,".D MHP``)*4``B2E``(4P/_TI*+__@#H""L4(/^``.`@(22E``(`J1`C!$$``@!` M""$D(0`!``$00Q````FDH/_^/`00`#P%$``DI0,H)(0&,`_@!%0`8#`A#!`4 MI"0$``&/OP`4)[T`&`/@``@`````)[W_Z`"@&"$D`@`!K[\`%`"`0"$48@!4 M`,!((13B`%(``````0`P(0$@$"&0Q```),8``3"%`'\PI?__$*`!58^_`!0P MC@"`$<``*@`````LH0`(%"``&@"@&"&0SP``)*7_^*!/``"0V``!,*7__Z!8 M``&0V0`"+*$`"*!9``*0R@`#)$(`"*!*__N0RP`$),8`"*!+__R0S/_]```` M`*!,__V0S?_^`````*!-__Z0SO__$"#_Z:!.__\`H!@A)*7__Q!@_]DPI?__ M`*`8(9#/```DI?__,*7__R3&``$D0@`!%&#_^:!/__\0`/_0D,0``)#$```L MH0`(%"``#B3&``$DI?_X,*7__RRA``B@1```H$0``:!$``*@1``#H$0`!*!$ M``6@1``&H$0`!Q`@__0D0@`(`*`8(22E__\08/^Z,*7__P"@&"$DI?__,*7_ M_Z!$```48/_[)$(``1``_[.0Q```$&(`!"0$``(0``!5)`0``B0$``(4Y`!2 M``````$`,"$!(!`AD,0``"3&``$PA0!_,*7__Q"@`/V/OP`4,)@`@!,``"H` M````+*$`"!0@`!H`H!@AD-D``"2E__BD60``D,H``3"E__^D2@`"D,L``BRA M``BD2P`$D,P``R1"`!"D3/_VD,T`!"3&``BD3?_XD,[__0````"D3O_ZD,__ M_@````"D3__\D-C__Q`@_^FD6/_^`*`8(22E__\08/_9,*7__P"@&"&0V0`` M)*7__S"E__\DQ@`!)$(``A1@__FD6?_^$`#_T)#$``"0Q```+*$`"!0@``XD MQ@`!)*7_^#"E__\LH0`(I$0``*1$``*D1``$I$0`!J1$``BD1``*I$0`#*1$ M``X0(/_T)$(`$`"@&"$DI?__$&#_NC"E__\`H!@A)*7__S"E__^D1```%&#_ M^R1"``(0`/^SD,0``!1D`%0`````%.(`4@`````!`#`A`2`0(93$```DQ@`" M,(4`?S"E__\0H`"HC[\`%#"*`(`10``J`````"RA``@4(``:`*`8(93+```D MI?_XH$L``)3,``(PI?__H$P``93-``0LH0`(H$T``I3.``8D0@`(H$[_^Y3/ M``@DQ@`0H$___)38__H`````H%C__939__P`````H%G__I3*__X0(/_IH$K_ M_P"@&"$DI?__$&#_V3"E__\`H!@AE,L``"2E__\PI?__),8``B1"``$48/_Y MH$O__Q``_]"4Q```E,0``"RA``@4(``.),8``B2E__@PI?__+*$`"*!$``"@ M1``!H$0``J!$``.@1``$H$0`!:!$``:@1``'$"#_]"1"``@`H!@A)*7__Q!@ M_[HPI?__`*`8(22E__\PI?__H$0``!1@__LD0@`!$`#_LY3$```49`!4```` M`!3D`%(``````0`P(0$@$"&4Q```),8``C"%`'\PI?__$*``4X^_`!0PC`"` M$8``*@`````LH0`(%"``&@"@&"&4S0``)*7_^*1-``"4S@`",*7__Z1.``*4 MSP`$+*$`"*1/``24V``&)$(`$*18__:4V0`(),8`$*19__B4RO_Z`````*1* M__J4R__\`````*1+__R4S/_^$"#_Z:1,__X`H!@A)*7__Q!@_]DPI?__`*`8 M(93-```DI?__,*7__R3&``(D0@`"%&#_^:1-__X0`/_0E,0``)3$```LH0`( M%"``#B3&``(DI?_X,*7__RRA``BD1```I$0``J1$``2D1``&I$0`"*1$``JD M1``,I$0`#A`@__0D0@`0`*`8(22E__\08/^Z,*7__P"@&"$DI?__,*7__Z1$ M```48/_[)$(``A``_[.4Q```/`00`#P%$``DI0-()(0&,`_@!%0`8#`A#!`4 MI"0$``&/OP`4)[T`&`/@``@`````)[W_R*^P`!@`@(`AK[\`')8.`'``H%`A M,<\`@@#`8"$5X``#`.!8(1```1,D`O__E@(`!``````L00`#$"```RQ!``(` M`%@A+$$``A`@``(```````!@(98#``(`````,&+_`!1``'4D`0$`$```:C!B M`/^6!``&C@@`#(X)`!".`P"$`(`0(0%`*"$00``2)(3__Y2X```DI0`"H'@` M`)!B`````````2((*Q`@``,`2`@K`$!((0!(""L0(``"``````!`0"$`@!`A M)(3__Q1`__`D8P`!K@@`#*X)`!`"`"`A`8`H(0P0#>``!``` M``"/I0`@#!`-8@%`("&6`@`&$```HX^_`!P\!!``/`40`"2E`V0/X`14)(0& M,`P0%*0D!``!$```FH^_`!PD`0`!$$'_E0`````D`0`"$$'_N0`````0`/_P M`````"0!`0`400"'`````!```'TP8@#_E@,`!HX(``P`8"`A`(`0(8X)`!`D MA/__$$``#P%`*"&4H@````````$B""L0(``#`$@(*P!`2"$`2`@K$"```@`` M````0$`A`(`0(22$__\40/_S)*4``HX&`(2N"``,K@D`$*^L`$"OJP!$KZ,` M$`%`("$D!0`"#!`.YR0'``&/I@!`CZ<`1*^B`"`"`"`A#!`.O0!`*"&/I0!` MCZ8`1`P0#><"`"`AC@4`A(^F`"`,$`Y)`@`@(98"``80``!`_@!%0DA`8P#!`4I"0$``$0```1C[\`'"0!``$00?^"`````"0!``(00?^P M`````!``__``````/`00`#P%$``DI0.,#^`$5"2$!C`,$!2D)`0``8^_`!R/ ML``8`^``"">]`#@GO?_8K[``$`"`@"&OOP`4KZ4`+)8.`'``P!@A,<\`@17@ M``,`````$```FB0"__^6`@`$`````"Q!``,0(``#+$$``@``."$L00`"$"`` M`P(`("$``!@A`@`@(0!@*"$,$`WG`.`P(98#``(`````,&+_`!1``$,D`0$` M$```.#!E`/^.!0"$E@8`!@P0#F<"`"`AE@8`!HX#`(0`!A0```(4`P!`*"$D M0O__``(4`(^D`"P0H``,``(4`P!`*"&0>```)$+__P`"%````A0#)&,``22$ M``(4H/_XI)C__I8&``8`````$```:@#`$"&6`P`&CZ4`+``#&$```QP```,< M`P!@,"&GHP`8#!`.9P(`("&&&0!RAZ,`&!,@``0`````CZ0`+`P0#6(`8"@A ME@(`!A```%B/OP`4/`00`#P%$``DI0.H#^`$5"2$!C`,$!2D)`0``1```$^/ MOP`4)`$``1"A_\<`````)`$``A"A_]\`````$`#_\``````D`0$`%$$`/``` M```0```R,&4`_PP0#J`"`"`AC@4`A``"-```!C0##!`.9P(`("&.!`"$CZ8` M+"0%``$,$!$T)`<``I8"``80```SC[\`%`P0#J`"`"`A``(<`(X%`(0``QP# M``(T```&-`.GHP`8#!`.9P(`("&&"`!RAZ,`&!$```0`````C@0`A`P0#6(` M8"@AC@0`A(^F`"PD!0`"#!`1-"0'``*6`@`&$```&H^_`!0\!!``/`40`"2E M`[P/X`14)(0&,`P0%*0D!``!$```$8^_`!0D`0`!$*'_S0`````D`0`"$*'_ MV0`````0`/_P`````#P$$``\!1``)*4#T`_@!%0DA`8P#!`4I"0$``&/OP`4 MC[``$`/@``@GO0`H`````">]_5@`@!@A)`[__Z^N`!P48``1K[\`%#P$$``/ MX`1L)(0%]!!```4`0!@AD$\````````5X``(`````).8@$``````$P```P`` M```0```C```0(2>#@$0\!1``)*4&`*.`@$`GI`(H#^`$O*^C`JB/HP*H)Z0" M-@_@!+P`8"@A)Z0"*`P0%,```"@A!$``$0!`("$GI0`F)`8"`@P0%-"OI``@ M)`$"`A1!``B/I``@/`00`"2$`_`GI0`F#^`$C"0&`@*OH``]_^BOOP`4#^`$/J^D M`!B/I``8#!`4X`````"/OP`4)[T`&`/@``@``````````"0"`^X````,$.`` M`P`````($!3D``````/@``@``!`A)`(#\`````P0X``#``````@0%.0````` M`^``"``````D`@/M````#!#@``,`````"!`4Y``````#X``(`````"0"`^P` M```,$.```P`````($!3D``````/@``@`````)`(#ZP````P0X``#``````@0 M%.0``````^``"``````D`@/[````#!#@``,`````"!`4Y``````#X``(```` M`"0"`^D````,`^``"``````\`1``K"(.;`/@``@D`O__/`,0`(QC#?0D`@/Y M`(,@(0````P4X``%`````#P!$`"L)`WT`^``"`!@$"$($!3D`````#P"$`", M0@WP``````""""L0(``"``````!`("$D`@/Y````#!3@__0`````/`$0`*PD M#?0#X``(```0(0`````````````````````\`A`!)$*A<#P!#\"L(@!`/`(0 M`21"P7@\`0_`K"(`1#P"$`$D0N&`/`$/P*PB`$@\`A``)$(&$#P!#\"L(@!, M/`(0`"1"#D`\`0_`K"(`4#P"$``D0@Q0/`$/P*PB`%0\`@!`)$)3H#P!#\"L M(@`4/`(`0"1"4]0\`0_`K"(`&#P"$``D0@Y@/`$/P*PB`!`\`A``)$(.;#P! M#\"L(@`F4Z(&-A;B=T(&]P96X@:6YP=70@9FEL90H`````:6UG7W-E96LZ('=I M97)D(&EM86=E('1Y<&4*`&EM9VQI8CH@EM<75Y?8$%"0T1%1D=(24I+3$U.3U!14E-455976%E: M>WQ]?G\````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M```````````````````````````````````````````````````````````` M0TA20TQ!4U,`````+VQI8B]C:')C;&%S Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Irix 3.3 Major_Enhancements_and_Compatibility Message-Id: <9517@odin.corp.sgi.com> References: <1990Jun26.010724.24482@calgary.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Just an FYI (and something I learned today ;-).. you can to "pcat file.z | ul -t dumb > file.txt" to get a file without the underlining stuff.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa04949; 27 Jun 90 1:34 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa04637; 27 Jun 90 0:47 EDT Received: from adm.brl.mil by VMB.BRL.MIL id aa04601; 27 Jun 90 0:29 EDT Received: from ucbvax.Berkeley.EDU by ADM.BRL.MIL id aa17303; 27 Jun 90 0:17 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA28157; Tue, 26 Jun 90 21:14:53 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 19:53:24 GMT From: "Bruce R. Holloway" Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Replacing the perspective matrix. Message-Id: <62902@sgi.sgi.com> References: <696@celia.UUCP>, <62846@sgi.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <62846@sgi.sgi.com>, tarolli@riva.esd.sgi.com (Gary Tarolli) writes: > In article <696@celia.UUCP>, keith@celia.UUCP (Keith Goldfarb) writes: > > > > What I want to do is to replace the current projection matrix > > (while in viewing mode) with one of my choosing with a call > > like loadmatrix(). Does such a call exist? > > The way to do this is to get into MPROJECTION mode, using > > mmode(MPROJECTION); > > and then use loadmatrix, and return to the previous mmode. In MVIEWING > mode, loadmatrix will load the model-view matrix, so that is why you must > get into MPROJECTION mode. This should work, but there may be a bug on > the GT/GTX machines (I think this was reported and fixed in the just > released 3.3 version). Gary is correct, except that the bug was replicated in the original Personal Iris graphics. I think this is the 4D/20 model only. Even though the PI & GT were fixed in 3.3 so that Gary's advice will work, there are still some limitations on the projection matrix you load. Basically, there can be no rotation in the projection matrix & any plane of projection must intersect only the z-axis. The matrices created by the GL calls which set the projection matrix all qualify. If you find it really necessary to load a more general projection matrix, you may report a bug. (I can't guarantee to fix it!) The GT graphics has been out for over two years & only one customer ever complained. He was satisfied with the 3.3 version. To the best of my knowledge, the new VGX graphics will handle a fully general projection matrix. Feel free to email me if you have further questions. There is a five- page technical report that is about 90% complete, but it was sort of written for internal consumption. Regards, bruceh   Received: from vmb.brl.mil by VMB.BRL.MIL id aa04991; 27 Jun 90 1:45 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ad04688; 27 Jun 90 1:02 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ad04655; 27 Jun 90 0:49 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa04898; 27 Jun 90 0:31 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA28460; Tue, 26 Jun 90 21:19:39 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 26 Jun 90 22:22:34 GMT From: Robert Skinner Organization: Silicon Graphics Inc., Entry Systems Division Subject: Re: Do somthing with a directory besides viewing Message-Id: <9528@odin.corp.sgi.com> References: <9006261127.AA07184@mcirps2.med.nyu.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9006261127.AA07184@mcirps2.med.nyu.edu>, root@MCIRPS2.MED.NYU.EDU writes: |> |> If I click on a directory below, I want to startup my appliation, not open the |> directory. Workspace does not seem to recognize these two matches. What am I |> doing bad ? |> |> TYPE SkullDirectory |> MATCH dircontains(".SkullDir"); |> # MATCH glob("D.Skulls"); |> CMD OPEN echo $LEADER |> ICON |> { |> include("SkullDir.fti"); |> } |> |> |> -- |> +-----------------------------------------------------------------------------+ |> | karron@nyu.edu Dan Karron | |> | . . . . . . . . . . . . . . New York University Medical Center | |> | 560 First Avenue \ \ Pager <1> (212) 397 9330 | |> | New York, New York 10016 \**\ <2> 10896 <3> | |> | (212) 340 5210 \**\__________________________________________ | |> +-----------------------------------------------------------------------------+ (Betsy may followup with a definitive response later, but for now:) The open/alt-open/etc.. rules for directories were originally interpreted like other files, but are now hard-coded in the workspace. Performance increased *dramatically* because directory rules are used so extensively. The performance increase was considered worth the loss of flexibility. (Page 2-10 of "Programming the IRIS WorkSpace" mentions this.) I believe the workspace-heads are working on relaxing this condition. Robert Skinner robert@sgi.com I saw a werewolf drinking a pina colada at Trader Vic's; an' his hair was perfect...   Received: from vmb.brl.mil by VMB.BRL.MIL id aa05438; 27 Jun 90 4:05 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa05236; 27 Jun 90 2:52 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa05215; 27 Jun 90 2:43 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa05090; 27 Jun 90 2:30 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA05872; Tue, 26 Jun 90 23:20:48 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 05:56:57 GMT From: Jeff Weinstein Organization: Silicon Graphics Inc. Subject: Re: G0 libraries Message-Id: <9572@odin.corp.sgi.com> References: <9006240711.16666@godzilla>, <3513@umbc3.UMBC.EDU>, <9555@odin.corp.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9555@odin.corp.sgi.com>, bean@putter.wpd.sgi.com (Bean Anderson) writes: > Programs performing dynamic loading will not necessarily "bomb" if they > weren't compiled -G 0. The main thing is that the dynamically loaded > modules were compiled -G 0; if not, the dynamically loaded module will > end up overlaying data space in the GP region. You could write a dynamic loader that does GP allocation and fixes up all references to GP data... --Jeff Jeff Weinstein - X Protocol Police Silicon Graphics, Inc., Entry Systems Division, Window Systems jsw@xhead.esd.sgi.com Any opinions expressed above are mine, not sgi's.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa05892; 27 Jun 90 5:45 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa05789; 27 Jun 90 5:24 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa05777; 27 Jun 90 5:11 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa05339; 27 Jun 90 5:00 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA15445; Wed, 27 Jun 90 01:55:25 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 04:00:30 GMT From: Trevor Paquette Organization: The University of Calgary Subject: XOR mode Message-Id: <1990Jun27.040030.17013@calgary.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Is there a cheap and dirty way to get XOR mode when doing a draw(), sboxf() or even bgnline()..endline()? I am porting an application that uses alot of the sunview 'PIX_SRC|PIX_DST' modes when writing to the screen. As far as I know about the sgi there is no 'built in' XOR mode for doing lines and/or block writes. I really don't want to write my own bres' algorithm for lines, checking for pixels along the way and XORing them myself, nor do I want to do that for each pixel in a block write. Can this be done easily? or should have have another go through the manuals and see what I missed. Trev ___________________________________________/No man is a failure who has friends Trevor Paquette ICBM:51'03"N/114'05"W|I accept the challange, body and soul, {ubc-cs,utai,alberta}!calgary!paquette|to seek the knowledge of the ones of old paquette@cpsc.ucalgary.ca | - engraved on the Kersa Blade of Esalon   Received: from vmb.brl.mil by VMB.BRL.MIL id ac09156; 27 Jun 90 9:31 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa08768; 27 Jun 90 9:20 EDT Received: from adm.brl.mil by VMB.BRL.MIL id aa08762; 27 Jun 90 9:12 EDT Received: from cunyvm.cuny.edu by ADM.BRL.MIL id aa23958; 27 Jun 90 9:05 EDT Received: from DDATHD21.BITNET by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 6914; Wed, 27 Jun 90 03:48:08 EDT Received: from BR2.THD.DA.D.EUROPE by DDATHD21.BITNET via GNET with RJE ; 27 Jun 90 09:49:35 Date: Wed, 27 Jun 90 09:47:54 +0200 (Central European Summer Time) From: Knobi der Rechnerschrat Subject: Named question To: info-iris@BRL.MIL X-VMS-To: X%"info-iris@brl.mil" Message-ID: <9006270905.aa23958@ADM.BRL.MIL> Hallo, I've just (thanks to the Examples) got the 3.3 named server up and running for our LAN. I have a question that I couldn't figure out from the docu- mentation (I fear the answer is no), but maybe some named guru can help me. On our LAN we have a couple of TCP/IP terminalservers. They are used to allow flexible terminal access to telnet-able hosts and (now the difficulties start) to provide service for computers not on the net. To do this, the computer is connected to one of the servers async ports and the server sets up a telnet service waiting on that port. To distinguis the ports (computers) the server waits on different tcp port-numbers (e.g 4711 instead of 23). The question: is there a way to generate a named entry that tells the requesting world that host xyz has IP address a.b.c.d and is waiting for telnet on port 4711 ? Any help is highly welcome Regards Martin Knoblauch TH-Darmstadt Physical Chemistry 1 Petersenstrasse 20 D-6100 Darmstadt, FRG BITNET:   Received: from vmb.brl.mil by VMB.BRL.MIL id aa10061; 27 Jun 90 10:00 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa09516; 27 Jun 90 9:48 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa09408; 27 Jun 90 9:37 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa06287; 27 Jun 90 9:31 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA00363; Wed, 27 Jun 90 06:28:01 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 12:29:49 GMT From: Gregory Fedor Organization: NASA/Lewis Research Center, Cleveland Subject: Re: convert sun images to IRIS image file format Message-Id: <1990Jun27.122949.29422@eagle.lerc.nasa.gov> References: <62933@sgi.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL A naive question: I've seen posted here programs in what appears to be a compiled format. I specifically am refering to the recent 'fromsun' program. What I'd like to know is what does one have to do to take the posted form and make it truly executable? I apologize if this has been hashed over before, but that was probably before I started looking at this group. Thanks in advance -- Hailing frequencies closed... Gregory Fedor (216) 433-8468 Sverdrup Technology smfedor@lerc01.lerc.nasa.gov NASA Lewis Research Center Cleveland, Ohio 44135 -------------------------------------------------------------------------------   Received: from vmb.brl.mil by VMB.BRL.MIL id aa10402; 27 Jun 90 10:15 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab10061; 27 Jun 90 10:04 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa09892; 27 Jun 90 9:50 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa06333; 27 Jun 90 9:38 EDT Received: from DDATHD21.BITNET by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 8585; Wed, 27 Jun 90 09:20:02 EDT Received: from BR2.THD.DA.D.EUROPE by DDATHD21.BITNET via GNET with RJE ; 27 Jun 90 15:05:25 Date: Wed, 27 Jun 90 15:03:44 +0200 (Central European Summer Time) From: Knobi der Rechnerschrat Subject: Problems with F77 from release 3.3 To: info-iris@BRL.MIL X-VMS-To: X%"info-iris@brl.mil" Message-ID: <9006270938.aa06333@VGR.BRL.MIL> Hallo, the following program gives a core dump when compiled/linked/run on a 3.3 4D/70GT. The program dumps during the second way through the loop. program test character*72 cline cline = ' 1 2 3 ' do 100 ii = 1 , 100 write (40,*) cline backspace (40) read (40,*) i,j,k write (*,*) i,j,k 100 continue end dbx shows that the problem is related to the backspace routines. Replacing backspace with rewind helps, but is NOT desirable. HELP !!! Regards Martin Knoblauch TH-Darmstadt Physical Chemistry 1 Petersenstrasse 20 D-6100 Darmstadt, FRG BITNET:   Received: from vmb.brl.mil by VMB.BRL.MIL id aa10647; 27 Jun 90 10:25 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab09516; 27 Jun 90 9:49 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ab09408; 27 Jun 90 9:37 EDT Received: from aero4.larc.nasa.gov by VGR.BRL.MIL id aa06291; 27 Jun 90 9:31 EDT Received: Wed, 27 Jun 90 08:14:44 EDT by aero4.larc.nasa.gov (5.52/5.6) Date: Wed, 27 Jun 90 08:14:44 EDT From: "Brent L. Bates AAD/TAB MS361 x42854" Message-Id: <9006271214.AA04951@aero4.larc.nasa.gov> To: van-bc!ubc-cs!alberta!arcsun.arc.ab.ca!calgary!cpsc!paquette@ucbvax.berkeley.edu Subject: Re: XOR mode Cc: info-iris@BRL.MIL Perhaps you need to use the alpha planes. (If you have them) -- Brent L. Bates NASA-Langley Research Center M.S. 361 Hampton, Virginia 23665-5225 (804) 864-2854 E-mail: blbates@aero4.larc.nasa.gov or blbates@aero2.larc.nasa.gov   Received: from vmb.brl.mil by VMB.BRL.MIL id aa10984; 27 Jun 90 10:51 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa10729; 27 Jun 90 10:40 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa10652; 27 Jun 90 10:25 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa06420; 27 Jun 90 10:01 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA02216; Wed, 27 Jun 90 06:55:57 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 13:37:43 GMT From: "Vasile R. Montan" Subject: Getting FrameMaker to display on an SGI Message-Id: <596@babcock.cerc.wvu.wvnet.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Has anyone had any success in getting FrameMaker to display on an SGI machine? I am running it on a Sun 4. I was getting the error: maker: using /usr01/frame/.makerinit maker: Cannot allocate colors. Aborting. However, someone suggested that I use adjacent colors for the background and foreground and using '-fg black -bg red' I would get past this error. There are still some problems in that the text on buttons does not show up, the tops of capital letters are cut of, and when I quit, all my X clients are killed. If anyone has had similar problems and knows a way around them, could you pass along your solution. Also if you never had a problem displaying FrameMaker on an SGI, could you send me a description of what you do. Thanks, -- Vasile   Received: from vmb.brl.mil by VMB.BRL.MIL id aa14225; 27 Jun 90 13:57 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa13690; 27 Jun 90 13:15 EDT Received: from adm.brl.mil by VMB.BRL.MIL id aa13556; 27 Jun 90 12:59 EDT Received: from ucbvax.Berkeley.EDU by ADM.BRL.MIL id aa29553; 27 Jun 90 12:47 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA12879; Wed, 27 Jun 90 09:31:55 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 15:45:50 GMT From: Toine Schreurs Organization: University of Utrecht, Dept. of Physics Subject: gifview revisited. Message-Id: <1256@ruunsa.fys.ruu.nl> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL There was a message about gifview some time ago. Is everybody satisfied with its behaviour? I'm not. You can't display more than one picture, and you can't move it once it's there. I made a simple viewer myself based on a prg downloaded from simtel. It's months ago. It can move pictures about. It needs 24 bitplanes. Anyone interested? -- Rob Hooft, hooft@chem.ruu.nl, hooft@hutruu54.bitnet   Received: from vmb.brl.mil by VMB.BRL.MIL id aa14444; 27 Jun 90 14:13 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab14225; 27 Jun 90 14:03 EDT Received: from adm.brl.mil by VMB.BRL.MIL id aa14166; 27 Jun 90 13:50 EDT Received: from ucbvax.Berkeley.EDU by ADM.BRL.MIL id aa00814; 27 Jun 90 13:40 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA16385; Wed, 27 Jun 90 10:21:19 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 16:55:41 GMT From: dave "who can do? ratmandu!" ratcliffe Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: 4Dgifts and 4sight server from 3.3 tapes. Message-Id: <9586@odin.corp.sgi.com> References: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article broderic@u2.rutgers.edu (Alfred J. Broderick) writes: > > A few weeks ago, Mark Callow mentioned that there were some > incompatibilities between 4Sight 3.3 and the ~4Dgifts/.4sight/*.ps > files. One of the problems was that the frame title gets > obscured by a black line. Unfortunately I did not save Mark's > message when it passed through the newsgroup. Was there a fix for > this problem? replace yer current version of ~4Dgifts/.4sight/window.ps with the following "fixed" file: --------------------------- SNIP-SNAP HERE ---------------------------------- systemdict begin % % This is used to draw the stripes on the title bar. Args are: % lslope,rslope slope [0-1] of edge of stripe on left and right side % step number of pixels between stripes % x, y lower left of area to be striped % w, h size of area to be striped % /fillslanted { % lslope rslope step x y w h => - % gsave 4 copy 0 1 0 setrgbcolor rectpath fill grestore 7 dict begin /h exch def /w exch def /y exch def /x exch def /step exch def /rslope exch step mul def /lslope exch step mul def 0 step h step sub { /y y step add def dup lslope mul x add y moveto dup x w add exch h exch sub rslope mul sub y lineto stroke pop } for end } def SGIWindow begin /KeyFocusColor 1.0 0.0 0.0 rgbcolor def /BannerColor 0.1 0.1 0.1 rgbcolor def /FrameFont /Times-BoldItalic findfont 14 scalefont def /IconFont /Times-BoldItalic findfont 14 scalefont def /Slantiness .3 def % how slanty are the stripes % (0 == not, 1 == most) /StripeSpacing 2 def % spacing of stripes on the % title bar (in pixels ) /PaintFrameLabel { % true/false => - (Paint frame text label) % % This should only be called by paintframelabel which does a gsave % and sets the canvas, color and font. % % set up clipping BannerButtonMargin BannerButtonWidth add TitleSideMargin add % t/f x1 dup FrameHeight BannerHeight sub 0.3753 sub % t/f x1 x1 y1 2 copy moveto dup 3 1 roll % t/f x1 y1 x1 y1 FrameWidth TitleSideMargin sub NoQuit? NoClose? or not { BannerButtonMargin BannerButtonWidth add sub } if % t/f x1 y1 x1 y1 x2 2 index max 1 index lineto % t/f x1 y1 x1 y1 0 BannerHeight rlineto BannerHeight add lineto % t/f x1 y1 0 BannerHeight neg rlineto clip % t/f x1 y1 3 -1 roll { % % Label already exists. % clear banner to make way for a different length title 0 exch FrameWidth BannerHeight bulbousrect PaintFocus } { pop } ifelse % x1 TitleSideMargin add TitleCapHeight dup add add % x1' CloseControl getcanvaslocation exch pop % x1' y currentfont fontheight 0.085 mul add % x1' y' XXX show paints too low moveto % - FrameLabel show } def /PaintFocus { gsave FrameCanvas setcanvas FrameFont setfont KeyFocus? {KeyFocusColor} {BannerColor} ifelse setcolor Slantiness 0 StripeSpacing xfc div % lslope rslope step CloseControl getcanvaslocation exch pop % ls rs st y1 0 Slantiness StripeSpacing xfc div % ... y1 ls' rs' st' NoStow? { BannerButtonMargin 4 index % ... y1 ls' rs' st' x y BannerButtonWidth TitleSideMargin dup add add TitleCapHeight dup add add % ... ls' rs' st' x y w } { BannerButtonMargin BannerButtonWidth add TitleSideMargin add 4 index % ... ls' rs' st' x y TitleCapHeight dup add } ifelse % ... ls' rs' st' x y w TitleCapHeight fillslanted % ls rs st y1 BannerButtonMargin BannerButtonWidth add TitleCapHeight dup add add TitleSideMargin dup dup add add add FrameLabel stringwidth pop add % ... y1 x1' exch % ... x1 y1 FrameWidth TitleSideMargin sub NoQuit? NoClose? or not { BannerButtonMargin BannerButtonWidth add sub } if % ... x1 y1 x2 2 index max % ... x1 y1 x2 2 index sub TitleCapHeight % ... x1 y1 w h fillslanted grestore } def end end   Received: from vmb.brl.mil by VMB.BRL.MIL id aa14632; 27 Jun 90 14:39 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa13002; 27 Jun 90 12:36 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa12896; 27 Jun 90 12:20 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa07117; 27 Jun 90 12:16 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA10949; Wed, 27 Jun 90 09:04:21 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 15:39:36 GMT From: Kurt Akeley Organization: sgi Subject: Re: XOR mode Message-Id: <9579@odin.corp.sgi.com> References: <1990Jun27.040030.17013@calgary.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <1990Jun27.040030.17013@calgary.uucp>, paquette@cpsc.ucalgary.ca (Trevor Paquette) writes: |> |> Is there a cheap and dirty way to get XOR mode when doing a draw(), |> sboxf() or even bgnline()..endline()? I am porting an application that |> uses alot of the sunview 'PIX_SRC|PIX_DST' modes when writing to the |> screen. As far as I know about the sgi there is no 'built in' XOR |> mode for doing lines and/or block writes. I really don't want to write |> my own bres' algorithm for lines, checking for pixels along the way |> and XORing them myself, nor do I want to do that for each pixel in a |> block write. |> Can this be done easily? or should have have another go through the |> manuals and see what I missed. |> Trev The GL command logicop() allows any of the 16 logical operations to be specified for all drawing commands (points, lines, polygons, surfaces, characters, etc.). Even the syntax PIX_SRC|PIX_DST is supported: logicop (LO_SRC | LO_DST); However, logicop is supported only by the newer hardware platforms. You can use the command getgdesc(GD_LOGICOP) to determine whether logicop is supported on a given machine. I believe that the Personal Iris with the RE2 chip, recently shipped GT/GTX systems with no alpha bitplanes, and VGX systems are the only machines that support logicop(). -- kurt   Received: from vmb.brl.mil by VMB.BRL.MIL id ab14632; 27 Jun 90 14:39 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa14563; 27 Jun 90 14:29 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa14552; 27 Jun 90 14:19 EDT Received: from ACF4.NYU.EDU by VGR.BRL.MIL id aa07422; 27 Jun 90 14:06 EDT Received: by acf4.NYU.EDU (5.61/1.34) id AA20734; Wed, 27 Jun 90 14:06:14 -0400 Date: Wed, 27 Jun 90 14:06:14 -0400 From: Dan Karron Message-Id: <9006271806.AA20734@acf4.NYU.EDU> To: info-iris@BRL.MIL, sgi!shinobu!odin!tristram@ucbvax.berkeley.edu Subject: Re: Panel Library from NASA Ames Cc: karron@mcirps2.med.nyu.edu, tristram@sgi.com If i read your signature correctly, you(David Tristram) are now working for sgi, and not nasa. Congratulations!!?? Does that mean that the sgi toolkit will look somwhat like the(your) nana panel lib. Are you or your former colleagues at nasa going to support the panel lib in irix 3.3 ? My comment on your dialog with Mike Gigante at MIT(where an X-bias can be detected) it that if he had to change the source code to get what he wanted, GREAT !. That is what source code is for. Did he send you his changes, and if they are valid features that should be added, did you include them in your source distribution. That is how these things evolve. I hope that the source to the sgi lib is not made priopriety for this very reason. dan.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa15129; 27 Jun 90 15:07 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa14748; 27 Jun 90 14:57 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa14693; 27 Jun 90 14:45 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa07508; 27 Jun 90 14:32 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA20545; Wed, 27 Jun 90 11:21:05 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 18:16:08 GMT From: Nicholas Tarnoff Organization: National Institute of Standards and Technology Subject: Unable to login if the login directory is NFS mounted ? Message-Id: <4846@marvin.cme.nist.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I have tried several times to manually add a /etc/passwd entry in order for a user to login to a home directory that is NFS mounted. The result is that the user has no console or toolchests. Perhaps there is a default configuration file for the console and toolchests which is no found when logging into NFS mounted home directories. Do you have any idea how to resolve this problem ? Thank you for your help ! -Nicholas ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Nicholas Tarnoff TELE: (301) 975-3464 NIST, Robot Systems Division FAX: (301) 990-9688 tarnoff@cme.nist.gov UUCP: uunet!cme-durer!tarnoff National Institute of Standards and Technology (formerly NBS) Bldg. 220 / Rm B127 Gaithersburg, MD 20899 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~   Received: from vmb.brl.mil by VMB.BRL.MIL id aa15490; 27 Jun 90 15:25 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ac15129; 27 Jun 90 15:14 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa15122; 27 Jun 90 15:07 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa07620; 27 Jun 90 15:00 EDT Received: from MCCLB0.MED.NYU.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 3731; Wed, 27 Jun 90 14:59:10 EDT Received: from mcirps2.med.nyu.edu by MCCLB0.MED.NYU.EDU; Wed, 27 Jun 90 13:14 EDT Received: by mcirps2.med.nyu.edu (5.52/890607.SGI) (for @mcclb0.med.nyu.edu:info-iris@brl.mil) id AA12618; Wed, 27 Jun 90 13:48:12 DSD Date: Wed, 27 Jun 90 13:48:12 DSD From: root%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: workspace: create a new file/directory in an empty dir. To: info-iris@BRL.MIL Message-id: <9006272048.AA12618@mcirps2.med.nyu.edu> X-Envelope-to: info-iris@brl.mil How do you make a new file or directory in an empty directory. With nothing to click on, the relevent pup lines are grayed out ! dan(somthing is wrong somplace) -- +-----------------------------------------------------------------------------+ | karron@nyu.edu Dan Karron | | . . . . . . . . . . . . . . New York University Medical Center | | 560 First Avenue \ \ Pager <1> (212) 397 9330 | | New York, New York 10016 \**\ <2> 10896 <3> | | (212) 340 5210 \**\__________________________________________ | +-----------------------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id aa16108; 27 Jun 90 15:52 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ab15490; 27 Jun 90 15:30 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa15411; 27 Jun 90 15:19 EDT Received: from cunyvm.cuny.edu by VGR.BRL.MIL id aa07644; 27 Jun 90 15:06 EDT Received: from MCCLB0.MED.NYU.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.2.2MX) with BSMTP id 3794; Wed, 27 Jun 90 15:04:37 EDT Received: from mcirps2.med.nyu.edu by MCCLB0.MED.NYU.EDU; Wed, 27 Jun 90 15:03 EDT Received: by mcirps2.med.nyu.edu (5.52/890607.SGI) (for @mcclb0.med.nyu.edu:info-iris@brl.mil) id AA13102; Wed, 27 Jun 90 15:37:05 DSD Date: Wed, 27 Jun 90 15:37:05 DSD From: root%mcirps2.med.nyu.edu@cunyvm.cuny.edu Subject: Workspace directory rule override secret To: betsy@sgi.com, info-iris@BRL.MIL Message-id: <9006272237.AA13102@mcirps2.med.nyu.edu> X-Envelope-to: info-iris@brl.mil, betsy@sgi.com The secret to bypassing the default behavior for directories is to use the SPECIALTYPE rule. I played around with the dumpster rule and learned this the hard way. This testing was done by doing a make clean on the /usr/lib/filtype directory, and stripping the rules in $HOME/.workspace to local/ and default/. Then I know what workspace knows and does not know. The this small subset of default default rules seems to cover most all file types. I played with the dumpster rule, because it is the only directory rule I could find in the sgi rules. Like it says below, i could not change the actions unless i cleared the default properties with the SPECIALTYPE rule. If I did somthing bad with the the SUPERTYPE or SPECIALFILE rule, it would not see the icon and think that the dumpster was a normal directory. Betsy told me not to use SUPERTYPE, but I did not understand what she had written. I wish she told me to use SPECIALFILE. That solved all problems!! .....from sgidefault.fti.....(edited version in /karron/.workspace/default) TYPE Dumpster MATCH dircontains(".dumpster"); LEGEND A dumy legendend is the dumpster # SUPERTYPE Directory # SUPERTYPE SpecialFile # SPECIALFILE # Note: the OPEN, ALTOPEN, and DROP rules are currently built in to the # workspace. Changing the placeholder rules here will not modify their # behavior. # CMD OPEN dirview $LEADER $REST # CMD ALTOPEN dirview -o `pwd` $LEADER # CMD DROP carefulmove $LEADER $REST $TARGET # CMD PRINT routeprint -g -t $LEADERTYPE $LEADER $REST #SUPERTYPE Directory #SUPERTYPE SpecialFile # all i want to do here so see what is happening. # putting back supertype only makes a mess. SPECIALFILE CMD OPEN echo CMD OPEN $LEADER $REST CMD ALTOPEN echo CMD ALTOPEN `pwd` $LEADER CMD DROP echo CMD DROP $LEADER $REST $TARGET CMD PRINT echo CMD PRINT $LEADERTYPE $LEADER $REST MENUCMD "E Dump" (echo $LEADER $REST) ICON { # Dumpster Shadow color(shadowcolor); ... PS: A null icon should not be a syntax error. You can note it, but for quick and dirty development, I don't care what the icon looks like, and a null icon should be allowed. The workspace windows should understand the wsh window scrolling function keys, page up, page down etc. The srolled position should be maintained between program invocations to be consistant with the window placement state being saved between program invocations. now... Back to the future!!! No More hacking muck! dan. -- +-----------------------------------------------------------------------------+ | karron@nyu.edu Dan Karron | | . . . . . . . . . . . . . . New York University Medical Center | | 560 First Avenue \ \ Pager <1> (212) 397 9330 | | New York, New York 10016 \**\ <2> 10896 <3> | | (212) 340 5210 \**\__________________________________________ | +-----------------------------------------------------------------------------+   Received: from vmb.brl.mil by VMB.BRL.MIL id aa18253; 27 Jun 90 18:29 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ab17413; 27 Jun 90 17:22 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa17382; 27 Jun 90 17:14 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa08100; 27 Jun 90 16:46 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA29443; Wed, 27 Jun 90 13:37:27 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 18:34:27 GMT From: Al Arthur Organization: NYIT Computer Graphics Laboratory Subject: Re: SGI warranties in presence of third party hardware Message-Id: <406@nyit.UUCP> References: <76170@aerospace.AERO.ORG> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <76170@aerospace.AERO.ORG> sinclair@aero.UUCP (William S. Sinclair) writes: >Has anyone ever had an instance of SGI voiding the warranty of their >equipment because of third-party SCSI equipment on the bus? This is a very good question. I would like to know what the policy is (in theory and practice), with regard to adding SCSI equipment to some of our Iris Equipment (ie. Exebyte tape etc.). How do SGI react to this equipment on systems that they maintain (service contract)? There would be a question as to whether the SGI scsi tape driver would work too... or would I have to start from scratch, and write my own? Any ideas... experiences... ? Thanks, Al -- Alex Arthur, System Programmer/UUCP Administrator New York Institute of Technology - Computer Graphics Laboratory Gerry House, Old Westbury, New York 11568 Phone:(516) 686-7644 UUCP: ...!{sbcs.sunysb.edu,philabs}!nyit!aca   Received: from vmb.brl.mil by VMB.BRL.MIL id aa18291; 27 Jun 90 18:39 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa18149; 27 Jun 90 18:18 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa18056; 27 Jun 90 18:05 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa08387; 27 Jun 90 18:01 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA04166; Wed, 27 Jun 90 14:47:45 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 19:46:46 GMT From: James Helman Organization: Stanford University Subject: Re: Getting FrameMaker to display on an SGI Message-Id: References: <596@babcock.cerc.wvu.wvnet.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL > However, someone suggested that I use adjacent colors for the > background and foreground and using '-fg black -bg red' I would get > past this error. Many X clients do not behave properly with static colormaps. Under IRIX 3.2's X server, the default visual was static rather then pseudo like most other common color X servers. Frame 1.3 can't deal with that, so you should start Xsgi with the -envm or -pseudo option to provide a different default visual, then you can use a white background. Under 3.3, the X server is started by default with the -envm option. But it's pointless. Like many old clients created back when X servers weren't too picky about illegal values, Frame 1.3 only works with the X server's "bug compatibility mode" turned on, e.g. xset bc. Under IRIX 3.3, Xsgi no longer supports this mode, so it's best to just wait until Frame fixes their problems. I'm hoping Frame 2.0X will be on better terms with Xsgi. Jim Helman Department of Applied Physics 6 Trillium Lane Stanford University San Carlos, CA 94070 (jim@thrush.stanford.edu) (415) 723-9127   Received: from vmb.brl.mil by VMB.BRL.MIL id aa18607; 27 Jun 90 20:13 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ac18525; 27 Jun 90 20:03 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id ac18511; 27 Jun 90 19:52 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa08612; 27 Jun 90 19:36 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA10526; Wed, 27 Jun 90 16:22:08 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 23:06:59 GMT From: Paul Haeberli Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Swell code fragment . . . . . . . Message-Id: <63004@sgi.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Hey! here's a little fragment of code that makes it easy to provide the user with a little moving hour-glass cursor to show percent done. paul haeberli paul@sgi.com /* * percent - * Indicate percent done on time consuming operations. * * Paul Haeberli - 1985 * * to use: * * . * . * winopen("name"); * . * . * percentdone(0.0); * for(y=0; y 99.9) { setcursor(oldcursor,0xfff,0xfff); curindex = -1; } else { index = percent*NCURSORS/100.0; if (index<0) index = 0; if (index>=NCURSORS) index = NCURSORS-1; if (index != curindex) { defcursor(20,curtab[index]); curorigin(20,7,7); setcursor(20,0xfff,0xfff); curindex = index; } } }   Received: from vmb.brl.mil by VMB.BRL.MIL id aa18664; 27 Jun 90 20:24 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa18525; 27 Jun 90 20:02 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa18503; 27 Jun 90 19:51 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa08593; 27 Jun 90 19:31 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA10482; Wed, 27 Jun 90 16:21:13 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 22:56:33 GMT From: Chris Schoeneman Organization: Silicon Graphics, Inc., Mountain View, CA Subject: gifview 2.1 Message-Id: <9641@odin.corp.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Hi there, for those of you interested in viewing GIF files, 'gifview 2.1' is available. (I know what you're saying: your 'gifview' doesn't have a version number. That's okay, you don't have 2.1) 'gifview 2.1' has the following features: * color map or RGB modes available (8,12,24 bits) * autodetection of available hardware * up to 8 color map pictures at once (with 12 bit planes) * 'unlimited' pictures in RGB mode * border can be turned on (since now it defaults to off) * position on screen can be specified on command line * images can be moved, stowed * images can be buffered when drawn for quicker redraws It's a little faster and the source is a lot easier to understand. If you're interested send me a note via email and send a copy out to you. I'll try to get this posted for anonymous FTP ASAP. I'll let you know when and where. Chris Schoeneman | I was neat, clean, shaved and sober, jindak@surfside.esd.sgi.com | and I didn't care who knew it. Silicon Graphics, Inc. | -Raymond Chandler Mountain View, CA | (The Big Sleep)   Received: from vmb.brl.mil by VMB.BRL.MIL id aa18721; 27 Jun 90 20:38 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab18664; 27 Jun 90 20:28 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa18656; 27 Jun 90 20:22 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa08696; 27 Jun 90 20:16 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA13307; Wed, 27 Jun 90 17:04:09 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 23:15:08 GMT From: martin Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Unable to login if the login directory is NFS mounted ? Message-Id: <9642@odin.corp.sgi.com> References: <4846@marvin.cme.nist.gov>, <9623@odin.corp.sgi.com> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9623@odin.corp.sgi.com> martinm@kmart.sgi.com (martin) writes: >In article <4846@marvin.cme.nist.gov> tarnoff@cme.nist.gov (Nicholas Tarnoff) writes: >> >>I have tried several times to manually add a /etc/passwd entry in >>order for a user to login to a home directory that is NFS mounted. >> >>The result is that the user has no console or toolchests. > >there is/was a bug in 3.2 where users with nfs mounted home directories >would lsoe things. the problem occured if there was anything in .login >or .cshrc that began with capital L set in a setenv statement, ie >if this line were in .cshrc " setenv LINES 40". this is one possibility >that i know of. hope it helps. > > >Martin McDonald >SGI > Life is unpredictable. > Eat dessert first. i did not post this very clearly. the problem is not with nfs, nor do files actually disappear. the problem is that NeWS could get confused, some of the symptoms being that console may not come up and you could end up with some of your toolchests missing. the scr number on this one was 8751, which was a duplicate of alpha bug 1970. after looking at these reports again, some of the cobwebs cleared from memory. after narrowing everything down on that particular call, nfs was not the problem, which probably won't help means this probably won't help you. sorry for posting an ambiguous response earlier. Martin McDonald SGI Life is unpredictable. Eat dessert first.   Received: from vmb.brl.mil by VMB.BRL.MIL id aa18848; 27 Jun 90 21:03 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa18781; 27 Jun 90 20:52 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa18739; 27 Jun 90 20:39 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa08732; 27 Jun 90 20:31 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA14272; Wed, 27 Jun 90 17:17:54 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 28 Jun 90 00:16:55 GMT From: Brent Thomas Corkum Organization: Rock Engineering, University of Toronto Subject: location of cursor in device coordinates Message-Id: <1990Jun27.201654.24688@jarvis.csri.toronto.edu> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL I am looking for a program that will tell me where my cursor is in device coordinates. Now this can be something that tells me the position as I move the cursor, or tells me the position when I hit a mouse button, it doesn't matter. Brent corkum@csri.toronto.edu   Received: from vmb.brl.mil by VMB.BRL.MIL id aa18923; 27 Jun 90 21:13 EDT Received: from vmb.brl.mil by VMB.brl.MIL id aa18463; 27 Jun 90 19:48 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa18443; 27 Jun 90 19:32 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa08567; 27 Jun 90 19:17 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA09986; Wed, 27 Jun 90 16:13:01 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 21:56:15 GMT From: martin Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Unable to login if the login directory is NFS mounted ? Message-Id: <9623@odin.corp.sgi.com> References: <4846@marvin.cme.nist.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <4846@marvin.cme.nist.gov> tarnoff@cme.nist.gov (Nicholas Tarnoff) writes: > > > > >I have tried several times to manually add a /etc/passwd entry in >order for a user to login to a home directory that is NFS mounted. > >The result is that the user has no console or toolchests. there is/was a bug in 3.2 where users with nfs mounted home directories would lsoe things. the problem occured if there was anything in .login or .cshrc that began with capital L set in a setenv statement, ie if this line were in .cshrc " setenv LINES 40". this is one possibility that i know of. hope it helps. Martin McDonald SGI Life is unpredictable. Eat dessert first.   Received: from vmb.brl.mil by VMB.BRL.MIL id ab18923; 27 Jun 90 21:13 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id ab18525; 27 Jun 90 20:02 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa18511; 27 Jun 90 19:52 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa08596; 27 Jun 90 19:32 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA10551; Wed, 27 Jun 90 16:22:34 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 23:11:28 GMT From: Vernon Schryver Organization: Silicon Graphics, Inc., Mountain View, CA Subject: Re: Unable to login if the login directory is NFS mounted ? Message-Id: <63005@sgi.sgi.com> References: <4846@marvin.cme.nist.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <4846@marvin.cme.nist.gov>, tarnoff@cme.nist.gov (Nicholas Tarnoff) writes: > I have tried several times to manually add a /etc/passwd entry in > order for a user to login to a home directory that is NFS mounted. > > The result is that the user has no console or toolchests. The most familiar cause is confusion among host name databases. Do the chests & console come up on the home machine (i.e. without NFS)? If not, then the problem is with NeWS and perhaps the user's user.ps If so, can you log into the account via telnet or rlogin? If not, then ensure that the home directories are properly mounted where the passwd database says they are--i.e. check `ls ~user` If so, other problems might involve having too many environment variables or something. Perhaps the .cshrc or .login are running a command that does not work on the local machine. Perhaps the user.ps is fancy and fiddling with the network. vjs   Received: from vmb.brl.mil by VMB.BRL.MIL id aa19216; 27 Jun 90 22:00 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa19124; 27 Jun 90 21:49 EDT Received: from adm.brl.mil by VMB.BRL.MIL id aa19101; 27 Jun 90 21:41 EDT Received: from ucbvax.Berkeley.EDU by ADM.BRL.MIL id aa10240; 27 Jun 90 21:34 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA18655; Wed, 27 Jun 90 18:22:50 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 28 Jun 90 00:39:31 GMT From: Mark Callow Organization: Silicon Graphics Inc., Entry Systems Division Subject: Re: 3.3 for the Personal IRIS Message-Id: <9651@odin.corp.sgi.com> References: <9006261147.AA14316@avelon.lerc.nasa.gov> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article <9006261147.AA14316@avelon.lerc.nasa.gov>, fsfacca@AVELON.LERC.NASA.GOV (Tony Facca) writes: |> |> > Does anybody know when IRIX 3.3 will be released for the Personal IRIS? |> > We just received a 340VGX with 3.3 installed and there are some nice |> > features I like to use on my PI. |> |> Yes, I'd like to know this as well. Also, would some one from SGI please |> answer the following: |> |> -- What rev of the OS is currently shipping for the 4D/25TG's? |> -- What rev of the OS is currently shipping for the 4D/3x0VGX's? |> I don't know what software releases manufacturing is installing on machines before shipping them. What I know is: Only Release 3.3 supports the VGX therefore the VGX is shipping with 3.3 The 4D1-3.3 software release supports all 4D hardware configurations therefore if you've got the tapes you can install 3.3 on any 4D you own. -- From the TARDIS of Mark Callow msc@ramoth.sgi.com, ...{ames,decwrl}!sgi!msc "There is much virtue in a window. It is to a human being as a frame is to a painting, as a proscenium to a play. It strongly defines its content."   Received: from vmb.brl.mil by VMB.BRL.MIL id aa19360; 27 Jun 90 22:25 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa19304; 27 Jun 90 22:14 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa19301; 27 Jun 90 22:08 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa09117; 27 Jun 90 22:02 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA20824; Wed, 27 Jun 90 18:55:40 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 28 Jun 90 00:52:10 GMT From: Mark Callow Organization: Silicon Graphics Inc., Entry Systems Division Subject: Re: 4Dgifts and 4sight server from 3.3 tapes. Message-Id: <9653@odin.corp.sgi.com> References: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL In article , broderic@u2.rutgers.edu (Alfred J. Broderick) writes: |> |> Hi! |> |> A few weeks ago, Mark Callow mentioned that there were some |> incompatibilities between 4Sight 3.3 and the ~4Dgifts/.4sight/*.ps |> files. One of the problems was that the frame title gets I DID NOT say that. The incompatibility is between the release 3.2.1 /usr/NeWS/lib/NeWS/sgiwin.ps and the 3.2 (and earlier) 4Dgifts/.4sight/window.ps subclass. A corrected window.ps was distributed with release 3.2.3. There is no problem in 3.3. ***IMPORTANT NOTICE*** If you are upgrading from 3.2 to 3.3 and you had copied ~4Dgifts/.4sight/window.ps to your own ~/.4sight, you *must* replace it with a copy of the window.ps distributed with 3.3. The corrected window.ps file for 3.2.[123] and 3.3 was posted to this group a few days ago by someone from our hotline. I originally posted this file several months ago, when I first discovered the problem. -- From the TARDIS of Mark Callow msc@ramoth.sgi.com, ...{ames,decwrl}!sgi!msc "There is much virtue in a window. It is to a human being as a frame is to a painting, as a proscenium to a play. It strongly defines its content."   Received: from vmb.brl.mil by VMB.BRL.MIL id aa20113; 28 Jun 90 1:08 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa19681; 27 Jun 90 23:45 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa19633; 27 Jun 90 23:30 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa09240; 27 Jun 90 23:17 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA25693; Wed, 27 Jun 90 20:09:02 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 28 Jun 90 03:02:48 GMT From: "K. Venkatesh Prasad" Organization: Rutgers Univ., New Brunswick, N.J. Subject: Making time measurements Message-Id: Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Dear sgi-netters, I am a newuser on the SGI machines and work on a 4D/50GT SGI machine. Would someone have any pointers to making accurate (millisecond) time measurements on the machine. What I need is the following. 0. I animate a figure or several different figures in time-sequence. 1. I need to have an observer "press" a mouse key to record observation instances ... the time intervals could be as short as a few milliseconds. Is there a way to do this on the SGI machine?? Many thanks in advance, Regards Prasad -- K. Venkatesh Prasad Machine Vision Group   Received: from vmb.brl.mil by VMB.BRL.MIL id aa20346; 28 Jun 90 2:26 EDT Received: from vmb.brl.mil by VMB.BRL.MIL id aa20329; 28 Jun 90 2:16 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id aa20316; 28 Jun 90 2:08 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa09553; 28 Jun 90 2:02 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA05499; Wed, 27 Jun 90 22:51:34 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 28 Jun 90 03:57:42 GMT From: Trevor Paquette Organization: The University of Calgary Subject: versions command + warning Message-Id: <1990Jun28.035742.2031@calgary.uucp> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL This is a word of warning to anyone who might be wanting to use the 'versions' command to 'remove' versions of software on an SGI. When using this command MAKE SURE THAT YOU ADD THE SUB_SYSTEM NAME OF THE SYSTEM YOU WANT TO REMOVE. Today I accidently hit return before typing in the name of the NFS subsystem that I wanted to remove. ie: what I wanted -> "versions remove nfs.sw.nfs" (something like that anyway) what I typed -> "versions remove nfs.sw.nfs". My hand brushed the return key accidently. When I realized what I did, ie the drive was taking WAY longer then it should have to delete JUST the NFS system, it was to late. I hit the 'delete' key (break on my system, 240GTX) and tried "ls" ...command not found.. "df" ..command not found.. (yikes) "ps" ..command not found..(don't look so good).. No problem I thought.. I'll just reload the 3.3 software.. "inst" .. command not found.. (starting to sweat a bit here).. "sync".. command not found.. "/etc/halt".. command not found.. Power down, wait 10 secs.. power up.. (this was to get to the "Install system software" menu on boot up(.. wait.. wait wait.. nothing.. a DEAD 240GTX.. I now have to get ahold of a boot tape and hope to God that I can salvage what was on the drives.. If I have to reformat I have lost alot of work. (yes I have backed up, but not for the past week). ***Why doesn't versions ask "ARE YOU SURE YOU WANT TO DO THIS?"?!?!?!?!?!*** It looks like it deleted all of the Unix commands and vmunix and god knows what else before I stopped it. Is there anyway I might be able to salvage what was on the disks once this happens??? ___________________________________________/No man is a failure who has friends Trevor Paquette ICBM:51'03"N/114'05"W|I accept the challange, body and soul, {ubc-cs,utai,alberta}!calgary!paquette|to seek the knowledge of the ones of old paquette@cpsc.ucalgary.ca | - engraved on the Kersa Blade of Esalon   Received: from vmb.brl.mil by VMB.BRL.MIL id ab12611; 29 Jun 90 8:40 EDT Received: from vmb.brl.mil by VMB.brl.MIL id ac12239; 29 Jun 90 8:30 EDT Received: from vgr.brl.mil by VMB.BRL.MIL id an11831; 29 Jun 90 8:09 EDT Received: from ucbvax.Berkeley.EDU by VGR.BRL.MIL id aa18391; 29 Jun 90 7:35 EDT Received: by ucbvax.Berkeley.EDU (5.63/1.41) id AA19279; Fri, 29 Jun 90 04:22:41 -0700 Received: from USENET by ucbvax.Berkeley.EDU with netnews for info-iris@brl.mil (info-iris@brl.mil) (contact usenet@ucbvax.Berkeley.EDU if you have questions) Date: 27 Jun 90 17:46:10 GMT From: Jim Tomlinson Organization: BoGART Graphics Project Subject: Best 8mm tapes? Message-Id: <938@voodoo.UUCP> Sender: info-iris-request@BRL.MIL To: info-iris@BRL.MIL Hope this hasn't been discussed. We just got a 4D/320S (gloat, gloat) with the 8mm cartridge drive. The question is, which manufacturer makes the best tape for digital use? Please e-mail; I'll summarize and post. Thanx. - jdt -- Jim Tomlinson P.O.Box 24346 \ BoGART Project M/S 6M-17 \ "Give me some of that Boeing Computer Services Seattle, WA 98124 \ voodoo thang!" - Colin James (206) 234-7741 ....uunet!bcstec!voodoo!jdt \