From guest Mon Aug 1 00:32:27 1994 To: info-performer@sgi.sgi.com Cc: aono@trl.ibm.co.jp Subject: pfNodeTravMask Date: Mon, 01 Aug 1994 16:32:13 +0900 From: Masaki Aono On page 147 of IRIS Performer Programming Guide (007-1680-020), there is an example of pfNodeTravMask. Specifically, it is written as pfNodeTravMask(root, PFSET_OR, PFTRAV_SET_FROM_CHILD, 0x0);. It seems that this is not a correct call, becasue C compiler does not accept this statement. Please let me know the right one. Regards, Masaki Aono Tokyo Research Laboratory From guest Mon Aug 1 01:38:42 1994 From: "Angus Henderson" Date: Mon, 1 Aug 1994 09:36:40 +0100 In-Reply-To: Simon Bennett "Realistic light points" (Jul 29, 8:04pm) References: X-Mailer: Z-Mail-SGI (3.0S.1026 26oct93 MediaMail) To: Simon Bennett , info-performer@sgi.sgi.com Subject: Re: Realistic light points Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 I was looking forward to see the answer to Simon Bennett's question - did someone reply in person and not copy info-performer.... ANgus From guest Mon Aug 1 05:24:57 1994 From: "Axel Sammet" Date: Mon, 1 Aug 1994 14:12:01 +0000 In-Reply-To: "Drew Hess" "swapbuffer timeouts" (Jul 26, 17:03) References: <9407261703.ZM5201@fusion.arc.nasa.gov> X-Face: %BF3$'?$)#p&@-.1*c72Cz/H4?x1i3iBWMbb1#mz6NKfMb}-Lv@@1dqU#3Yb@PTVm@ui#]@ 2]L/vOL[CZ{*FbsNs=gbZ8S-oJb7|0;.Os~S3nDH@CUZQ]J*XyqX ZSz.YX}_v_j>@YVxsy5Tbpg$+_\@;T8jTK9xueFT}Q5x1*"+gz0 X-Mailer: Z-Mail-SGI (3.0S.1026 26oct93 MediaMail) To: "Drew Hess" , info-performer@sgi.sgi.com Subject: Re: swapbuffer timeouts Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Hi Drew, Unfortunately you didn't mention which Performer Version you're using! We experienced this behavior in Performer 1.1 (IRIX 5.1.1.3). Switching to IRIX 5.2 and Performer 1.2 eliminated all off these problems. Regards Axel -- ______________________________________________________________________ Axel Sammet, SE | Tel: +49 511 9017216 Silicon Graphics | Fax: +49 511 6138115 Ahrensburger Strasse 3 | VM #: 59191 M/S: IHA-314 30659 Hannover | E-Mail: axels@hannover.sgi.com ______________________________________________________________________ From guest Mon Aug 1 05:56:13 1994 (5.65c/IDA-1.4.4 for ); Mon, 1 Aug 1994 08:56:01 -0400 Date: Mon, 1 Aug 94 08:55:45 EDT From: fyock@capella.tsc.gtefsd.com (Christina Fyock) To: info-performer@sgi.sgi.com, metivier@sanders.com Subject: Re: port to 1.2 I had the same VClock error and this was the fix I got from SGI. It worked for me so I hope it works for others. After your winopen() call, you need to put in a pfInitVClock(0) call. This call sets the vertical retrace counter to ticks and enables CPU interrupts from the graphics pipeline. I hope this helps! Christina Fyock From guest Mon Aug 1 06:08:46 1994 (5.67a/IDA-1.5 for info-performer@sgi.com); Mon, 1 Aug 1994 08:10:31 -0500 Date: Mon, 1 Aug 1994 08:10:31 -0500 From: Milt Fulghum To: Simon Bennett Subject: Re: Realistic lightpoints again... Cc: info-performer@sgi.sgi.com Reply-To: fulghum@vss.fsi.com > > iv) This is actually an important issue? i.e. is it important > to know the individual sizes/power of individual > lightpoints or groups of lightpoints... > >Has anybody done any work or research in this area at all? I'd like >to hear about any sucess or failure stories at all. (Modelling and/or >rendering) We have always thought that lightpoint intensity was a big enough issue to include it in our standand feature list. It seems that for flight simulation applications the power and size of lightpoints provide important visual cues. Based on this some of our products have included lightpoint perspective growth to give realistic lightpoint presentations. Milton Fulghum ----------------------------------------------------------------------------- Milton L. Fulghum PHONE: (314) 925-8576 FlightSafety International, Inc. FAX: (314) 925-8444 2590 North Highway 94 E-Mail: fulghum@vss.fsi.com Saint Charles, MO 63301-0037 From guest Mon Aug 1 09:09:37 1994 From: "ted jordan" Date: Mon, 1 Aug 1994 12:09:04 -0400 X-Face: j;kCV?eS+]ChRWl;VY/JLQw'A6B/%I(w,LT|3G9"H+!pjDW>%3s#S|I\}Fbs;ZH_d]GTE26H"ZAW_;)h6J,&\kX/%fmo'O?lmwvZ<9vaKm=(TuGBlm0/w-a~E/iwtvSGJw3/7g|IaG9D0npq,l>vkt:zK:rm,u%>O%aw=0&]@HgHWG[K=MO3 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: info-performer@sgi.sgi.com Subject: Swap > 1GB causing errors in 1.1 Performer Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Did this get fixed in 1.2? tj -- theodore p mugabe jordan silicon graphics, inc detroit, michigan email: tjordan@sgi.com regional service engineer fax: 810-615-2157 "From what we get, we can make a living; what we give, makes a life", "Days of Grace", A. Ashe Monthly Spelling Aid: usage: how to use it; not useage From guest Mon Aug 1 10:35:35 1994 Date: Mon, 1 Aug 1994 10:35:37 -0700 From: hitchner@netcom.com (Lew Hitchner) To: info-performer@sgi.sgi.com Subject: Re: pfNodeTravMask Masaki Aono writes: >> On page 147 of IRIS Performer Programming Guide (007-1680-020), >> there is an example of pfNodeTravMask. >> Specifically, it is written as >> pfNodeTravMask(root, PFSET_OR, PFTRAV_SET_FROM_CHILD, >> 0x0);. >> It seems that this is not a correct call, becasue >> C compiler does not accept this statement. >> Please let me know the right one. >> Regards, >> Masaki Aono >> Tokyo Research Laboratory Masaki's question raises a question with me. In the errant example from page 147 of the Programming Guide, the call to pfNodeTravMask uses the `setMode' parameter value of PFTRAV_SET_FROM_CHILD. According to the manuals I have (Prog. Guide and Reference Pages) the only valid values for that parameter are PFTRAV_SELF and PFTRAV_DESCEND. I can't find any mention of PFTRAV_SET_FROM_CHILD except in the example on page 147. Is PFTRAV_SET_FROM_CHILD a valid value, and, if so, where can we find an explanation of it. Lew Hitchner Virtual Reality and Visual Simulation Consultant Mountain View, CA From guest Mon Aug 1 11:33:49 1994 From: fraser@portola.com (Glen D. Fraser) id AA02594 for info-performer@sgi.com; Mon, 1 Aug 94 11:33:57 PDT Subject: Stencilling and transparency To: info-performer@sgi.sgi.com Date: Mon, 1 Aug 1994 11:33:57 -0700 (PDT) X-Mailer: ELM [version 2.4 PL21] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1490 I posted a question last week about transparent objects and the use of stencil planes. I've received no comments about it, and am still experiencing the problem. This is the setting: - I am working with multiple stencil planes in a Performer 1.2 application; - I have some transparent objects (objects with material alpha < 1.0) in the same scenes as the stencil ones; - I also have some objects with textures containing alpha components, - I have antialiasing turned on (using multisampling on our single-pipe Onyx RE2 with 2 RMs). Whenever I use stencil planes and multisampling (using mssize(8, 32, 8), for example), the objects with material alpha no longer appear. However, textures with alpha components still seem to render correctly. If I turn off stencilling, then the transparent objects return, but, of course, my stencil stuff is all messed up. Should I be able to display transparent objects and make use of one or more stencil planes at the same time? Do I need a greater pixel depth than "medium"? Could someone offer some suggestions? Thanks. Glen. =========================================================== Glen D. Fraser Head, Software and Telepresence Research, Inc. System Development 320 Gabarda Way Phone: (415)854-4420 Portola Valley, CA 94028 Fax: (415)854-3141 USA E-Mail: fraser@telepresence.com =========================================================== From guest Mon Aug 1 13:08:45 1994 Date: Mon, 1 Aug 1994 16:08:31 -0400 From: zeus@cs.UMD.EDU (Hillel Steinberg) To: info-performer@sgi.sgi.com Subject: Performer 1.2 woes... Hello Performers, I am using Performer 1.2 and have accomplished much with exception to the following three problems: 1. I need to draw labels on my objects as they float through the virtual world so I use a pfNodeTravFuncs function callback for post drawing on the objects. To make sure that the label appears I use a: zfunction(ZF_ALWAYS); zwritemask (0x0); within the callback. But, the labels aren't entirely drawn because other objects drawn behind (such as terrain) cover up the label. Is there any way to make sure that NOTHING overwrites the label without using the OVERLAY plane? In other packages, there were ways to give priorities to each object to control the order in which they were rendered. I have not spotted such calls in Performer 1.2, and do not want to resort to ordering pfAddChild's to force an order (because I have objects being introduce/removed dynamically which will complicate things). 2. I have a 22 Meg RGB file of a map that has words which are easily read with an 'ipaste'. If I overlay this texture onto a huge terrain mesh, and then move the camera in and out from the textured terrain, the map's detail is not sufficient to read the words on it no matter where the camera is positioned (I assume this is exercising the min/mag filters which I have tried many combinations). Why is the detail of the texture perfect with an 'ipaste', but not when it is mapped in Performer 1.2 to a mesh and viewed??? I would gladly sacrafice computation power for clarity in the texture. 3. On exit, I do a pfExit()/exit() which hangs for about 10 seconds. The Performer 1.2 FAQ does mention a bug similar to this, but: a. I run in PFPHASE_FREE mode anyway, so I shouldn't be affected. b. I see no inactive window/draw process because I am running single CPU. How do I prevent this delay EVERY time I exit the application? If you have any ideas on ANY of the three problems I have mentioned, give me a buzz at address: zeus@cs.umd.edu Thanks in advance, - Hillel - From guest Mon Aug 1 15:21:19 1994 ; Mon, 1 Aug 1994 18:20:59 -0400 id AA01686; Mon, 1 Aug 1994 15:22:44 PST Organization: MultiGen Inc X-Charset: MACINTOSH X-Umcp-To: INFO PERFORMER From: Marcus To: INFO PERFORMER Date: Mon, 1 Aug 1994 14:13:02 PST Subject: Re: >Loadflt problems? Reply to: RE>>Loadflt problems? >Hi Mark > >I have the same problem when loading a lot of objects with a big >amount of geometry. To force the loader to crash you can use simple.c >and add an unlimited loop around the "LoadFile" call. Use esprit.flt >for example. >The workaround LoadFltMode(PFFLT_COMBINELODS,FALSE), which was >suggested by Mike Weiblein is working; the limit is higher now. > >Regards, >Ekki Klaerner. Yes, turning off combine lod's does help. I discovered this early on and told Mike Wieblen to try it out as a work around. combineLODs() in the Flight loader makes potentially many calls to pfDelete() and pfRemoveChild() which appear to aggrevate (create?) the problem. By turning off combine lod's, you get an extended lease on life ... Regards, Marcus Barnes, Member Technical Staff MultiGen Inc., 1884 The Alameda, San Jose CA, 95126 PH: (408) 261 4110 FX: (408) 247 4329 EMAIL: multigen!marcus@uunet.UU.NET From guest Tue Aug 2 01:09:34 1994 From: desmond@iss.nus.sg (Desmond Hii Toh Onn) (931110.SGI//ident-1.0) id AA04539; Tue, 2 Aug 94 16:10:59 +0800 Subject: async file IO To: info-performer@sgi.sgi.com Date: Tue, 2 Aug 1994 16:10:59 +0800 (SST) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 794 Greetings, I'm having some problem with async file io. I cannot write anything to it. There *must* be something I did not do. Thanks for the advice in advance. below is the code desmond /*________________________________________________________________*/ #include #include #include #include #include main() { pfFile *oFile; int Fstatus; oFile = pfOpenFile("test.open",O_WRONLY|O_CREAT|O_APPEND,0644); Fstatus = pfWriteFile(oFile,"first line\n",11); printf("status of file %d\n",Fstatus); /* returns 0 !!! */ Fstatus = pfGetFileStatus(oFile,PFRTF_STATUS); printf("status of file %d\n",Fstatus); /* return 1 */ Fstatus = pfCloseFile(oFile); printf("status of file %d\n",Fstatus); /* return 0 */ } From guest Tue Aug 2 03:04:42 1994 Tue, 2 Aug 94 12:02:50 GMT From: "Ekkehard Klaerner" Organization: AITEC GmbH & Co KG To: info-performer@sgi.sgi.com Date: Tue, 2 Aug 1994 12:02:19 GMT+1 Subject: swapbuffer timing Priority: normal X-Mailer: PMail v3.0 (R1a) Hi, I know, that in the previous beta-versions of PF1.2 there was a problem with the swapbuffer timing. If the system is semi-loaded and working in FLOAT mode with 30Hz, it tends to switch between 30 and 60 Hz. This happens just in the draw-process while the update-rate of the application-process is constant ! I understand, that in the production release of PF1.2 the problem has been fixed, but sometimes I have the impression, that it still happens. Has someone else experienced this same problem ? Ekki. AITEC GmbH & CO Informationstechnologie KG Ekkehard Klaerner Alter Hellweg 50 D-44379 Dortmund Tel.: 0231/96465-45 Fax.: 0231/9646598 EMail: ek@aitec.de From guest Tue Aug 2 09:42:46 1994 Date: Tue, 2 Aug 94 09:44:42 PDT From: barker@cs.nps.navy.mil (Randall Barker) To: info-performer@sgi.sgi.com Subject: fltloader and a texture I have been porting an application from performer 1.1 to 1.2. I am having a problem with the multigen loader. We are using version 13 I am told. The problem is that all of the textures are loading fine except for one. I have run the old version (of the program) to make certain it was still loading the same texture and it is. I have viewed the texture to make certain it was not messed up and the texture looks fine (using imgview). I have watched the texteures as they are eached displayed at start up of the program. The texture is not displayed correctly there either. Some times also, it looks like other texture (ie smoke, fire etc.) are being used in its place. I have turned off all of the loader options. And have tried may differnt combinatons. I am running the application on a RE2 with IRIX 5.2. Any sugestions, comments are welcome, thanks -rb From guest Tue Aug 2 10:14:09 1994 To: info-performer@sgi.sgi.com Subject: error in pfuSmoke utility Date: Tue, 02 Aug 1994 13:13:37 -0400 From: Todd Metivier There seems to be a slight error in the pfuSmoke utility program smoke.c . When you try and make a pfuSmoke of type PFUSMOKE_MISSLE, the application will crash. This is because on line 178 in smoke.c in the initPuff routine, pfScaleVec3 tries to use a NULL pointer to a pfVec3 called rv. The vector rv is only initialized if smoke->type is not PFUSMOKE_MISSLE as shown below : /* Add random velocity modifier */ if(smoke->type != PFUSMOKE_MISSLE) { rv = RAND_VEC3; pfAddScaledVec3(puff->direction, smoke->direction, .15f, rv); } else PFCOPY_VEC3(puff->direction, smoke->direction); pfScaleVec3(offset, smoke->expansion * smoke->radius * .05f, rv); PFADD_VEC3(puff->origin, puff->origin, offset); This can be fixed by placing the last two lines of code inside the if statement. These lines of code simply offset the puff origins by random amounts. This presumable does not need to happen for missile smoke. fix : /* Add random velocity modifier */ if(smoke->type != PFUSMOKE_MISSLE) { rv = RAND_VEC3; pfAddScaledVec3(puff->direction, smoke->direction, .15f, rv); pfScaleVec3(offset, smoke->expansion * smoke->radius * .05f, rv); PFADD_VEC3(puff->origin, puff->origin, offset); } else PFCOPY_VEC3(puff->direction, smoke->direction); Todd Metivier | ARPA: metivier@sanders.com Lockheed Sanders | UUCP: ...!uunet!sanders.com!metivier PTP2-A001 | 65 River Road | Voice: (603) 885-9026 Hudson, NH 03051 | Fax: (603) 885-0631 From guest Tue Aug 2 11:22:22 1994 From: "Drew Hess" Date: Tue, 2 Aug 1994 11:22:03 -0700 In-Reply-To: "Ekkehard Klaerner" "swapbuffer timing" (Aug 2, 12:02pm) References: <658B5BA3F7A@aitec.de> X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: "Ekkehard Klaerner" Subject: Re: swapbuffer timing Cc: info-performer@sgi.sgi.com Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Yes, I had this problem last week. I added some compute-intensive code to the application process of my Performer app and it caused occasional swapbuffer timeouts. I haven't had the problem lately though; sounds like it is load-dependent. -dwh- dhess@vision.arc.nasa.gov From guest Tue Aug 2 13:27:02 1994 From: "Andy Bushnell" Date: Tue, 2 Aug 1994 13:08:09 -0700 In-Reply-To: ag "this from performer line - want to reply?" (Aug 1, 9:44am) References: <9408011644.AA00839@cory> X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: nova.unix.portal.com!portal!sgi.com!info-performer Subject: Performer Light points Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Replying to this info request... > More on this lightpoints issue of mine... > > Assuming that Performer can be "rigged" to handle the display of > "realistic" lightpoints or that a future (near future???) version of > Performer can handle them itself... I've got a few database > modelling queries... Multigen v.14 doesn't seem to support the notion > of giving lightpoints any size and/or power attributes... Does > anybody know if: > > i) Yes this is true. > > ii) No this isn't true.. you do it like this... > > iii) No Multigen doesn't do it.. but XYZ modeller does... > > iv) This is actually an important issue? i.e. is it important > to know the individual sizes/power of individual > lightpoints or groups of lightpoints... > > Has anybody done any work or research in this area at all? I'd like > to hear about any sucess or failure stories at all. (Modelling and/or > rendering) > > Once again.. much thanks for any and all assistance rendered.. > > > > +--------------------------------------------------------------------------------+ > Simon Bennett simonb@wormald.com.au > Wormald Technology Advanced Systems Engineering Ph: +61 2 981 0611 (x512) > Let me try and answer your questions without appearing to be ..tooooo.. biased. First of all I should point out (if you didn't know allready) that our company (Coryphaeus Software Inc) is kinduv in competition with MultiGen(r). 1. Im not altogether sure about this one (having never even seen v14) *but* our 3D Vis-Sim modeler (Designer's Workbench - DWB) *does* support light point size and many other light point/light string attributes. (more below). 2. With the risk of appearing cheeky I'd offer the solution that DWB is how you do it !! 3. Ah ha, now for the juicy bit ! DWB fully supports the following light point features & attributes... o Light point directionality can be set OMNI, UNI or BI-DIRECTIONAL. o Light points/strings can be assigned overall/front or back colors (depending on the light directionality). o Light point type flag can be set (one of Runway,Taxiway,City,VASI, Centerline etc etc). o Light point shape hint flag can be set (straight,curved,random etc) o Light point size (in pixels) can be set. o Light Point real-world size (light point diameter). This is used by DWB to automatically build polygonal lights that could be used under a Switch Node (LOD) to switch them in at near distances. o Light beam with & height for UNI-DIRECTIONAL lights can be set. o Intensity & Intensity variance along a light string can be set. o Calligraphic priority (0->1) indicates the relative importance of different light strings. Calligraphic lights are (usually) a limited resource so this field indicates the most important ones. o The direction vector can be viewed & manipulated. o A "suppress last light" flag is available that indicates that the end of string #1 is also the start point of string #2. 4.This is really based on the desired fidelity of the end-simulation. I would think that size of light points is important so that the overall "weighting" of the different light strings when compared to each other can be guaged. Currently different "power" levels could be frigged in Performer by setting different Light point size values in DWB. In addition to the attributes listed above DWB has 9 tools to lay down straight,curved or random light points/strings. DWB has its own *PUBLIC DOMAIN* binary format for which a Performer loader is available and provided in source. The subset of DWB's light point attributes that Performer supports (in its pfLightPoint Node) are all fully supported in our loader. If you want more info on DWB, our Performer support or our other Visual Simulation products (terrain,IG tools) please dont hesitate to contact me. Andy Bushnell, Director of Engineering, Coryphaeus Software Inc, 985 University Av. Suite#31, Los Gatos, CA 95030 Tel : 408 395 4537 Email : ab@cory.coryphaeus.com From guest Wed Aug 3 00:27:00 1994 To: davec@cory.coryphaeus.com Cc: info-performer@sgi.sgi.com Subject: Re: pfLoadTexture vs pfTexImage In-Reply-To: Your message of "Thu, 28 Jul 94 11:42:57 -0800." <9407281842.AA06704@cory> Date: Wed, 03 Aug 94 00:26:13 -0700 From: Jim Helman There aren't many things that could cause texture paging. Either you are somehow creating multiple pfTextures for the same file or the internal format is indeed different. My guess is that some other part of the loader code is doing a pfGetTexName to decide whether it should load the texture file. ** == change. ** if (alpha) pfTexFormat(pfTex,PFTEX_INTERNAL_FORMAT,PFTEX_RGBA_4); ** else pfTexFormat(pfTex,PFTEX_INTERNAL_FORMAT,PFTEX_RGB_5); output = decodeTexture(path); /* create the texture... I initially had 1 for the nr parameter value but I thought that maybe that gave it a depth of 2 (ie 0 & 1) so I now set it to zero but it makes no difference to the problem */ pfTexImage(tex,output,nComponents,nWidth,nHeight,0) ** pfTexName(tex, path); rgds, -jim helman jimh@surreal.asd.sgi.com 415/390-1151 From guest Wed Aug 3 00:45:45 1994 To: "ted jordan" Cc: info-performer@sgi.sgi.com Subject: Re: Swap > 1GB causing errors in 1.1 Performer In-Reply-To: Your message of "Mon, 01 Aug 94 12:09:04 EDT." <9408011209.ZM3203@macgyver.detroit.sgi.com> Date: Wed, 03 Aug 94 00:36:05 -0700 From: Jim Helman > Did this get fixed in 1.2? Performer 1.2 should work with >1GB of swap space, although I don't have a machine of sufficient stature to test it out on. rgds, -jim helman jimh@surreal.asd.sgi.com 415/390-1151 From guest Wed Aug 3 03:19:04 1994 From: "pierre" Date: Wed, 3 Aug 1994 10:47:58 -0600 X-Mailer: Z-Mail-SGI (3.0S.1023 23oct93 MediaMail) To: info-performer@sgi.sgi.com Subject: refresh problem Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Thus anybody can help me please ? When i use a callback in the culling process to draw a GL object on my scene, I have some refresh trouble on the cloud (the old vision of my object is not clear), when my object are on the ground it's perfect.( My config is Onyx RE2 Irix 5.2 and Performer 1.2) I haven't see in the documentation that it exist some parameters to tell if the cloud are refresh or not during the cycle. Have you an idea ? Thanks in advance frank PS : last info you see the same effect when the oject isn't include in the performer application. I explain, you take an icon an you put that on top of you performer application. When this icon is over the ground the refresh is ok when it is over the cloud the refresh isn't ok. From guest Wed Aug 3 04:34:26 1994 From: "Robert Rossow" Date: Wed, 3 Aug 1994 07:35:27 -0400 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: info-performer@sgi.sgi.com Subject: please remove me Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Please remove me from the list for noe... thanks, Rob rossow@gsc.gsi.com From guest Wed Aug 3 06:29:03 1994 Date: Wed, 3 Aug 94 09:28:54 EDT From: louis@speech.iar.nrc.ca (Louis Bolduc) To: info-performer@sgi.sgi.com Subject: pfGSetLineWidth broken? Ok, I'm no Performer wizard, but this sounds like pretty basic stuff... I can't seem to be able to get lines wider than 1 pixel. Here's a piece of code that should give me lines five pixels wide, but invariably gives me 1 pixel wide lines... what's wrong ? My system is an Indigo Elan running IRIX 5.2 and Performer 1.2... #include #include #include #include #include #include static pfVec3 coords [] = { { 0.0F, -30.0F, 0.0F }, { 0.0F, 30.0F, 0.0F }, { -30.0F, 0.0F, 0.0F }, { 30.0F, 0.0F, 0.0F }, { 0.0F, 0.0F, -30.0F }, { 0.0F, 0.0F, 30.0F } }; static pfVec4 colors [] = { { 1.0F, 1.0F, 1.0F, 1.0F } }; static pfGeode *make_lines ( void ) { void *arena = pfGetSharedArena (); pfGeoSet *gset = pfNewGSet ( arena ); pfGeoState *gstate = pfNewGState ( arena ); pfGeode *lines = pfNewGeode (); pfGSetAttr ( gset, PFGS_COORD3, PFGS_PER_VERTEX, coords, NULL ); pfGSetAttr ( gset, PFGS_COLOR4, PFGS_OVERALL, colors, NULL ); pfGSetPrimType ( gset, PFGS_LINES ); pfGSetNumPrims ( gset, 3 ); pfGSetLineWidth ( gset, 5.0F ); pfGSetGState ( gset, gstate ); pfAddGSet ( lines, gset ); return lines; } static void OpenPipeline ( pfPipe *p ) { foreground (); prefposition ( 100, 500, 100, 500 ); winopen ( "IRIS Performer" ); pfInitGfx ( p ); pfApplyMtl ( pfNewMtl ( pfGetSharedArena () ) ); pfApplyLModel ( pfNewLModel ( pfGetSharedArena () ) ); } main () { float t = 0.0f; pfScene *scene; pfPipe *p; pfChannel *chan; pfInit (); pfMultiprocess ( PFMP_DEFAULT ); pfConfig (); scene = pfNewScene (); pfAddChild ( scene, make_lines () ); pfAddChild ( scene, pfNewLSource () ); p = pfGetPipe ( 0 ); pfInitPipe ( p, OpenPipeline ); chan = pfNewChan ( p ); pfChanScene ( chan, scene ); pfChanNearFar ( chan, 1.0f, 1000.0f ); pfChanFOV ( chan, 45.0f, 0.0f ); pfInitClock ( 0.0f ); while ( t < 20.0f ) { float s, c; pfCoord view; pfSync (); t = pfGetTime (); pfSinCos ( 45.0f * t, &s, &c ); pfSetVec3 ( view.hpr, 45.0f * t, -10.0f, 0 ); pfSetVec3 ( view.xyz, 100.0f * s, -110.0f * c, 30.0f ); pfChanView ( chan, view.xyz, view.hpr ); pfFrame (); } pfExit (); return 0; } From guest Wed Aug 3 08:49:01 1994 From: "pierre VERCRUYSSE" Date: Wed, 3 Aug 1994 16:50:02 -0600 X-Mailer: Z-Mail-SGI (3.0S.1023 23oct93 MediaMail) To: info-performer@sgi.sgi.com Subject: (Fwd) refresh problem Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Sorry i have done a mistach, it isn't on a Onyx, the configuration is Irix 4.0.5 Performer 1.0 on Crimson VGX --- Forwarded mail from "pierre" To: info-performer@sgi.sgi.com Thus anybody can help me please ? When i use a callback in the culling process to draw a GL object on my scene, I have some refresh trouble on the cloud (the old vision of my object is not clear), when my object are on the ground it's perfect.( My config is Onyx RE2 Irix 5.2 and Performer 1.2) I haven't see in the documentation that it exist some parameters to tell if the cloud are refresh or not during the cycle. Have you an idea ? Thanks in advance frank PS : last info you see the same effect when the oject isn't include in the performer application. I explain, you take an icon an you put that on top of you performer application. When this icon is over the ground the refresh is ok when it is over the cloud the refresh isn't ok. --- End of forwarded mail from "pierre" -- -- -- Pierre Vercruysse pierre@rennes.sgi.com vmail : 58781 Tel (33) 99-23-12-80 Fax (33) 99-23-18-95 -- . _ _ _ _ Silicon Graphics France | | | | | | | | | office RENNES -- | |- |- |- |- Espace Performance | | |_ |\ |\ |_ 35769 Saint Gregoire From guest Wed Aug 3 09:56:50 1994 From: "Shahram Shiri" Date: Wed, 3 Aug 1994 12:56:37 -0400 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: info-performer@sgi.sgi.com Subject: Problem: generating .ptu format in Performer Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 I am new to Performer and trying to generate a terrain model from two images. One image is the terrain elevation and the other is the image of the location. Chapter 9 of Performer Programming Guide exlpains the .ptu format for this purpose, but there is no examples of the file in the book or Performer source codes. Based on the book's description of the .ptu file I generated the following terrain.ptu file : terrain 2 5 5 1 1 terrain.bw 255 255 land.bw Also, I followed the suggestions on the Makefile and regenerated the perfly application by linking the ImageVision libraries and uncommented the GETTILE for libpfsgi such that it reads the ilgettile.c file. When I run the following command perfly terrain.ptu I get these messages: Performer Info (9): All 1 processors available on this machine. Performer Warning (9): ilgettile: PTU requires perfly linked with IL. This requires Performer Warning (9): that the ImageVision Library be installed, that the definition of Performer Warning (9): GETTILE in the Makefile for libpfsgi be changed to ilgettile.c, Performer Warning (9): and that you uncomment the lines defining 'LIBIL' and 'LIBCIL' Performer Warning (9): in the Makefile for perfly Performer Warning (9): WARNING: Default Texture being used Performer Info (9): ITILE(0,0): <0.000 0.000> to <-1.000 -1.000> at <1 1> It repeates itself for each node. At the end, just a white plane appears on the Performer perfly window. My image file is 255x255 and terrain data is another image file by the same dimensions. I will apprecaite any suggestions and help on getting this problem resolved. -Thanks, Ron -- ------------------------------------------------------------------------- |Shahram Shiri/ ron@leaf.gsfc.nasa.gov Tel: 301-286-9985 | |NCCS/Scientific Visualization Studio Fax: 301-286-1634 | |NASA/Goddard Space Flight Center, Hughes/STX | |Greenbelt, Maryland 20771 | ------------------------------------------------------------------------- From guest Wed Aug 3 13:22:12 1994 Date: Wed, 3 Aug 1994 13:20:11 -0700 From: Phillip Polinsky To: info-performer@sgi.sgi.com Subject: more pfuSmoke utility Trying to get missile smoke to look like missile smoke. In my function to update the missile position, I set the smoke origin to the missile position using the pfuSmokeOrigin function. i.e. pfSetVec3(origin, missile_position_x, missile_position_y, missile_position_z); pfuSmokeOrigin(smoke, origin, radius); In Application function smoke type is set to missile. i.e. pfSmokeType(smoke, PFUSMOKE_MISSLE); pfuSmokeMode(PFUSMOKE_START); When I run the program there is about half a dozen puffs in front of the missile, and nothing behind it. Definitely not to impressive. Tried hacking smoke.c with limited success. Can the pfuSmoke utility do better??? Does anybody have some sample code that generates reasonable looking missile smoke that can run realtime??? Thanks in advance!!! Phillip Polinsky | Voice: (619) 939-0500 Naval Air Warfare Center | pill@archimedes.chinalake.navy.mil China Lake, Ca 93555 | From guest Wed Aug 3 21:30:47 1994 To: louis@speech.iar.nrc.ca (Louis Bolduc) Cc: info-performer@sgi.sgi.com Subject: Re: pfGSetLineWidth broken? In-Reply-To: Your message of "Wed, 03 Aug 94 09:28:54 EDT." <9408031328.AA07096@speech.iar.nrc.ca> Date: Wed, 03 Aug 94 21:29:55 -0700 From: Jim Helman Your program runs fine on VGX and RE. But GL's linewidthf() (which Performer uses to set line widths) appears to be broken in GL on Indy and Elan. For now, adding the following stub to your application should resolve the problem: void linewidthf(float x) { linewidth((int)(x+0.5f)); } rgds, -jim helman jimh@surreal.asd.sgi.com 415/390-1151 From guest Thu Aug 4 13:48:11 1994 From: halliday@BanffCentre.AB.CA (Sean Halliday) Subject: PFGS_QUADS To: info-performer@sgi.sgi.com Date: Thu, 4 Aug 1994 14:49:07 -0700 (MDT) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 532 Performer 1.2 ONYX RE2 IRIX 5.2 I have a quad that looks like: ------------ \ / \ / \ / \____/ I am texture mapping it with each corner of the quad as each corner of the texture map. What I get is the quad split into two triangles that have texture maps that do not line up. Is there a way to turn off the triangulation? -- Sean Halliday Computer Graphics Software Developer New Media Research, Banff Centre for the Arts. halliday@BanffCentre.AB.CA From guest Fri Aug 5 00:38:42 1994 Fri, 5 Aug 94 9:36:46 GMT To: info-performer@sgi.sgi.com From: "Andreas Roessler" Date: Fri, 5 Aug 1994 09:36:30 GMT Subject: unsubscribe Reply-To: andreas.roessler@iao.fhg.de Priority: normal X-Mailer: WinPMail v1.0 (R2) unsubscribe ****************************************************** *** Andreas Roessler *** *** *** *** FhG-IAO *** *** Nobelstr. 12c *** *** 70569 Stuttgart *** *** Germany *** *** *** *** Tel. 0711-970-2087 Fax. 0711-970-2299 *** *** *** *** EMail: Andreas.Roessler@iao.fhg.de *** ****************************************************** From guest Fri Aug 5 04:40:55 1994 Date: Fri, 5 Aug 94 07:31:57 -0400 From: steve@pith.umenfa.maine.edu (Stephen Shaler) To: info-performer@sgi.sgi.com Subject: unsubscribe Stephen Shaler Forest Products Laboratory University of Maine steve@pith.umenfa.maine.edu From guest Fri Aug 5 05:16:40 1994 Date: Fri, 5 Aug 94 08:17:15 EDT From: htree@gdstech.grumman.com (Hong-Yi Ip) To: info-performer@sgi.sgi.com Subject: unsubscribe unsubscribe From guest Fri Aug 5 08:44:15 1994 id AA21142; Fri, 5 Aug 1994 10:43:51 -0500 From: fuc@beloit.edu (Ching-hawn Fu) Subject: unsubscribe To: info-performer@sgi.sgi.com Date: Fri, 5 Aug 1994 10:43:51 +22310826 (CDT) -------------------------- Ching-Hawn Fu Beloit Corporation CAE/CAM Dept. Beloit, WI 53511 Email: fuc@beloit.edu Tel: (608)364-7107 Fax: (608)364-7013 -------------------------- X-Mailer: ELM [version 2.4 PL20] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 24 please unsubscribe me ! From guest Fri Aug 5 15:08:28 1994 Date: Fri, 5 Aug 1994 15:06:43 -0700 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: info-performer@sgi.sgi.com From: jan@archimedes.chinalake.navy.mil (Jan A. Barglowski) Subject: Vertex normals in pfuBuilder? Hello all! I'm playing with pfuBuilder to take an array of points and create a terrain, but I'd really like to get some vertex normals instead of the face normals. Anyone have a trick to do that before I start hacking builder.c? Many thanks! jan. Jan Anthony Barglowski Phone: (619) 927-1057 Computer Dude Internet: jan@archimedes.nwc.navy.mil Visualization Lab Packet: SYSOP@WA6YBN.#SOCA.CA.US.NA Naval Air Warfare Center kc6uth@kc6uth.ampr.org -Weapons Division [44.17.2.6] From guest Sun Aug 7 18:16:05 1994 (5.67b8/IDA-1.5 for info-performer@sgi.com); Mon, 8 Aug 1994 11:15:50 +1000 (5.65c/IDA-1.5 for ); Mon, 8 Aug 1994 11:09:27 +1000 (5.65c/IDA-1.4.4 for ); Mon, 8 Aug 1994 11:08:23 +1000 Date: Mon, 8 Aug 1994 11:12:21 +1000 (EST) From: Simon Bennett Subject: Re: more pfuSmoke utility To: info-performer@sgi.sgi.com In-Reply-To: <199408032020.NAA27553@sorcerer.chinalake.navy.mil> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Wed, 3 Aug 1994, Phillip Polinsky wrote: > Trying to get missile smoke to look like missile smoke. > In my function to update the missile position, I set the smoke origin to the > missile position using the pfuSmokeOrigin function. > i.e. > pfSmokeType(smoke, PFUSMOKE_MISSLE); > pfuSmokeMode(PFUSMOKE_START); > When I run the program there is about half a dozen puffs in front of the > missile, and nothing behind it. Definitely not to impressive. > Tried hacking smoke.c with limited success. > Can the pfuSmoke utility do better??? > Does anybody have some sample code that generates reasonable looking missile > smoke that can run realtime??? I've also been "hacking" smoke.c to generate all sorts of effects. With some more sucessful than others.. Apart from missle trails is there other Performer compatible special effects code hanging around anywhere? I went to an "EasyScene" demo held by Coryaphaeus a couple of months back. Had some very nice fire and smoke in it which they said they got off SGI... The Marconi Oil Rig demos have excellent fire and smoke... but I doubt that this code is available... +--------------------------------------------------------------------------------+ Simon Bennett simonb@wormald.com.au Wormald Technology Advanced Systems Engineering Ph: +61 2 981 0611 (x512) "Business meetings are really very useful... For one thing they clearly demonstrate just how many people the company can do *without*" From guest Mon Aug 8 03:47:08 1994 From: "Angus Henderson" Date: Mon, 8 Aug 1994 11:44:39 +0100 In-Reply-To: Simon Bennett "Re: more pfuSmoke utility" (Aug 8, 11:12am) References: X-Mailer: Z-Mail-SGI (3.0S.1026 26oct93 MediaMail) To: Simon Bennett , info-performer@sgi.sgi.com Subject: Re: more pfuSmoke utility Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 I must learn to resist the temptation to answer this type of e-mail. I helped do the smoke that coryphaeus use. It is in GL as a postDraw call to performer. One of the people that helped me write it was also the person who did the MArconi smoke - ANgus Dorbie (now) of DIVISION. I am not keen to sent it out at the moment becayse it has some short cuts that reduce performance. We intend to make a CD of some of the stuff we do here at The Reality Centre and we would also like to put some of it on the net, however I dont have access to a permanent slot on a public machine. I noticed Allan Schaffers mail on "http://www.sgi.com/tech/Performer.html" can someone out there advise ? Angus Henderson The Reality Centre From guest Mon Aug 8 07:26:14 1994 From: jamison@jib.engin.umich.edu (James Joseph Korotney) Subject: unsubscribe To: info-performer@sgi.sgi.com Date: Mon, 8 Aug 1994 10:27:44 -0500 (EDT) X-Mailer: ELM [version 2.4 PL21] Content-Type: text Content-Length: 779 unsubscribe me please thank you ____________________________________________________________________________ And as always I would remind you to please phrase your responses in the form of a question. -Alex ---------------------------------------------------------------------------- James J. Korotney University of Michigan __ /| jamison@vr1.engin.umich.edu Virtual Reality Laboratory \ o.0' 2600 Draper Rd. =(___)= UofM: (313) 763-7798 rm. 236D NAME building U fax: (313) 936-8820 Ann Arbor, MI 48109-2145 Ackkk! home: (313) 668-8766 ---------------------------------------------------------------------------- From guest Mon Aug 8 07:36:07 1994 (5.65c/IDA-1.4.4 for <@datasrv.co.il:info-performer@sgi.com>); Mon, 8 Aug 1994 17:36:10 +0300 From: alon@bvr.co.il (Alon Rosenfeld 18 Hatzedef street Jaffa ) Date: Mon, 8 Aug 1994 17:33:51 +0000 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: info-performer@sgi.sgi.com Subject: earth sky drawing Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 hi, I would like to know how Performer constructs its sky, horizon and ground polygons as well as why is it implemented this way. Thanks Alon From guest Mon Aug 8 07:55:45 1994 From: "Michael Jones" Date: Mon, 8 Aug 1994 07:55:04 -0700 In-Reply-To: "Angus Henderson" "Re: more pfuSmoke utility" (Aug 8, 11:44am) References: <9408081144.ZM11702@death.reading.sgi.com> X-Mailer: Z-Mail (3.2a.627 27jun94 MediaMail) To: "Angus Henderson" , Simon Bennett , info-performer@sgi.sgi.com Subject: Re: more pfuSmoke utility Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii We'll be establishing a public ftp site for IRIS Performer goodies, such as Angus' new smoke (which is very hot), the new IRIS Inventor loader, and other materials. We'll post to the mailing list when it's all set up, and in the interim, please let me know if you have anything to add to the collection. Thanks, Michael Jones -- Be seeing you, Phone:415.390.1455 Fax:415.390.2658 M/S:8U-590 Michael T. Jones Silicon Graphics, Advanced Graphics Division mtj@sgi.com 2011 N. Shoreline Blvd., Mtn. View, CA 94039-7311 From guest Mon Aug 8 09:08:30 1994 Date: Mon, 8 Aug 1994 09:08:32 -0700 From: hitchner@netcom.com (Lew Hitchner) To: info-performer@sgi.sgi.com, mtj@babar Subject: Re: more pfuSmoke utility >> We'll be establishing a public ftp site for IRIS Performer >> goodies, such as Angus' new smoke (which is very hot), the >> new IRIS Inventor loader, and other materials. >> We'll post to the mailing list when it's all set up, and in >> the interim, please let me know if you have anything to add >> to the collection. >> Thanks, >> Michael Jones Could you please include an example that shows how to used indexed colors with pfColorTable's. No one has replied to my request for help with and examples of using indexed colors that I posted a couple of weeks ago (during the busy SIGGRAPH partying, er I mean, work week). Lew Hitchner Virtual Reality and Visual Simulation Consultant Mountain View, CA From guest Mon Aug 8 15:31:14 1994 From: giraffe.asd.sgi.com!sgi.sgi.com!portal.unix.portal.com!portal!cory.coryphaeus.com!davec (David Cooper) Subject: Re: more pfuSmoke utility To: nova.unix.portal.com!portal!sgi.com!info-performer Date: Mon, 8 Aug 1994 14:26:22 -0800 (PDT) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1314 > >I've also been "hacking" smoke.c to generate all sorts of effects. With >some more sucessful than others.. Apart from missle trails is there >other Performer compatible special effects code hanging around anywhere? >I went to an "EasyScene" demo held by Coryaphaeus a couple of months >back. Had some very nice fire and smoke in it which they said they got >off SGI... > >The Marconi Oil Rig demos have excellent fire and smoke... but I doubt >that this code is available... The fire you saw demo'd by Coryphaeus is available as a dwb file. If you have pfdwb.c v2.2 ( which is unlikely coz only v2.1 is shipped with the current release of Performer) then this can be loaded directly and you will have animated fire, exactly as you saw in the demo. Smoke will be available shortly. The only code you need to write will be to call LoadFile and perhaps the addition of a matrix if you want the flames somewhere else than the origin. Recompile with pfdwb.c (v2.2) and thats all. pfdwb.c and flames.dwb are public domain and are available freely to anyone who wants them. Email me at davec@coryphaeus.com -- Dave Cooper Coryphaeus Software, Inc. davec@coryphaeus.com 985 University Ave., Suite 31, Los Gatos, CA 95030 Tel: 408-395-4537 Fax: 408-395-6351 From guest Tue Aug 9 03:56:38 1994 From: "Angus Henderson" Date: Tue, 9 Aug 1994 11:54:11 +0100 In-Reply-To: alon@bvr.co.il (Alon Rosenfeld 18 Hatzedef street Jaffa ) "earth sky drawing" (Aug 8, 5:33pm) References: <9408081733.ZM8216@rea1.bvr.co.il> X-Mailer: Z-Mail-SGI (3.0S.1026 26oct93 MediaMail) To: alon@bvr.co.il (Alon Rosenfeld 18 Hatzedef street Jaffa ), info-performer@sgi.sgi.com Subject: Re: earth sky drawing Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Oh no, I'm at it again.... Here is an idea from Mr. Clive Harding of Maritime Dynamics - I don't think he's on the info-performer list - that I found to work very well. The idea is based on the fact that it is good to use a PFES_TAG aka PFCL_MSDEPTH clear. That is because it clears the z-buffer for free with the assumption that all pixels in the display are covered once with polygon geometry. The most efficient way to use your ONYX is to use tag clear and ensure that the database approximates to a box around the eyepoint. this always results in the lowest depth complexity at the worst case viewpoint. Assuming the ground is covered by database geometry at or around z = 0.0 then to use a tag clear the sky must be drawn with a large polygon at the far clip plane extending from "high up" to z = 0.0 ( or minimum ground height ). This must be the intention of PFES_SKY_CLEAR except it uses a fast clear instead of a tag clear. I don't think Performer 1.2 has pfESkyMode option that does this - it would be a "PFES_SKY_TAG" i suppose. What Clive suggested was an upturned bowl that sits around the eyepoint and represents the sky. the bowl extends down to intersect with the distant terrain and has the advantage that it can be given a nice horizon and sky texture to provide more pitch & heading cues when you can't see the ground. The "skyBox" has travFuncs that turn off lighting and fog etc. and it is set to be haze coloured at the horizon so you can't see the join. It has a DCS that keeps it at the eyepoint and is attached to the scene to pick up h,p,r. It also has an updateTimeOfDay function very similar to that in the perfly application. Is this of any interest to you info-performer ? ANgus From guest Tue Aug 9 05:43:53 1994 From: plevy@tabor.mitre.org (Paul Levy) Subject: Re: earth sky drawing To: info-performer@sgi.sgi.com Date: Tue, 9 Aug 94 8:43:31 EDT X-Mailer: ELM [version 2.3 PL11] Another technique for drawing the sky is to use one polygon with an alpha cloud texture on it. The polygon is located directly above the eyepoint, sloping slightly towards the horizon, and extending to the far clipping plane. To add a bit of realism, when the eyepoint moves forward, we scroll the cloud texture backwards and move the polygon with the eyepoint. It looks as though you are flying through scattered clouds. Since you can see through the clouds to the EarthSky, changing the earthsky colors and adding lightpoints for stars can result in a very realistic looking sky at any time of the day. -- -------------------------------------------------------------------------- Paul Levy - The MITRE Corporation - McLean, Virginia 703-883-7421 -------------------------------------------------------------------------- From guest Wed Aug 10 08:24:40 1994 Wed, 10 Aug 94 17:22:17 GMT From: "Ekkehard Klaerner" Organization: AITEC GmbH & Co KG To: info-performer@sgi.sgi.com Date: Wed, 10 Aug 1994 17:21:53 GMT+1 Subject: projection-matrix Priority: normal X-Mailer: Pegasus Mail v3.1 (R1a) Hi, Is it possible to define a projection-matrix of my own ? If so, do I have to modify GL-matrices in the DrawCallback ? What about performance in this context ? Ekki. AITEC GmbH & CO Informationstechnologie KG Ekkehard Klaerner Alter Hellweg 50 D-44379 Dortmund GERMANY Tel.: 0231/96465-45 Fax.: 0231/9646598 EMail: ek@aitec.de From guest Wed Aug 10 08:23:23 1994 Wed, 10 Aug 94 17:21:14 GMT From: "Ekkehard Klaerner" Organization: AITEC GmbH & Co KG To: info-performer@sgi.sgi.com Date: Wed, 10 Aug 1994 17:20:56 GMT+1 Subject: pfTranslate Priority: normal X-Mailer: Pegasus Mail v3.1 (R1a) Hi , Is the use of pfTranslate and pfRotate in nodeDrawCallbacks faster than pfSCS, if we just need translation and rotation ? Ekki. AITEC GmbH & CO Informationstechnologie KG Ekkehard Klaerner Alter Hellweg 50 D-44379 Dortmund GERMANY Tel.: 0231/96465-45 Fax.: 0231/9646598 EMail: ek@aitec.de From guest Wed Aug 10 11:00:03 1994 From: "ted jordan" Date: Wed, 10 Aug 1994 13:59:11 -0400 In-Reply-To: "Ekkehard Klaerner" "pfTranslate" (Aug 10, 5:20pm) References: <71E0A743789@aitec.de> X-Face: j;kCV?eS+]ChRWl;VY/JLQw'A6B/%I(w,LT|3G9"H+!pjDW>%3s#S|I\}Fbs;ZH_d]GTE26H"ZAW_;)h6J,&\kX/%fmo'O?lmwvZ<9vaKm=(TuGBlm0/w-a~E/iwtvSGJw3/7g|IaG9D0npq,l>vkt:zK:rm,u%>O%aw=0&]@HgHWG[K=MO3 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: info-performer@sgi.sgi.com, dougr@orl.mmc.com Subject: Please add this cust to the info perf list Cc: tjordan@macgyver.detroit.sgi.com Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Doug Richter Martin Marietta dougr@orl.mmc.com tj -- theodore p mugabe jordan silicon graphics, inc detroit, michigan email: tjordan@sgi.com regional service engineer fax: 810-615-2157 "From what we get, we can make a living; what we give, makes a life", "Days of Grace", A. Ashe Monthly Spelling Aid: escalate: take to higher level; not esculate From guest Wed Aug 10 13:19:59 1994 From: aaron@qbert.dseg.ti.com (Aaron Hightower) Subject: projection-matrix & pfTranslate To: info-performer@sgi.sgi.com Date: Wed, 10 Aug 1994 15:16:14 -0500 (CDT) Reply-To: aaron@dseg.ti.com X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 1842 --- From: "Ekkehard Klaerner" >Is the use of pfTranslate and pfRotate in nodeDrawCallbacks faster >than pfSCS, if we just need translation and rotation ? If you use a pfDCS, the bounding volume of the child tree is recomputed automatically for you. Otherwise, you'll have to keep track of it yourself. If Performer keeps track of whether your matrix if affine or not (and it could rather easily) it shouldn't be any faster to do the translation and rotation in a callback. Check page 286 of the programming guide for details on affine matrices. The Performer team will have to verify if you would gain anything at all from using a draw callback for this, but I'll go out on a limb and guess, no. >Is it possible to define a projection-matrix of my own ? If you want to define your own projection matrix, you should be able to do so without having to explicitly create the matrix. See the man page on pfFrustum. If you want to position the eyepoint via a matrix, use pfChanViewMatrix. >If so, do I have to modify GL-matrices in the DrawCallback ? The man page for pfFrustum details the effect of the viewing Performer pfFrustum on the GL matrices. >What about performance in this context ? Again, I'm guessing, but I doubt there is a speedup. You would most likely be reimplementing some of Performer unnecessarily. Hope this helps, _ _ | | *-------------------------------------------* | | __| |___ | Aaron.Hightower@dseg.ti.com 214.575.6759 | __| |___ \ / | Simulation & Planning Technology | \ / \_ / | 6620 Chase Oaks Blvd M/S 8518 | \_ / \ ( | Plano TX 75023 | \ ( \/ *-------------------------------------------* \/ From guest Wed Aug 10 14:26:51 1994 Date: Wed, 10 Aug 1994 21:26:32 GMT From: "Carlo L. Tiana" To: info-performer@sgi.sgi.com Subject: network installation Cc: chad@vision.arc.nasa.gov We run performer on a number of machines (Onyx and Indigo 2) and would like to avoid having a full distribution of Performer on each machine, but rather would like to mount it from the server for all the machines. Unfortunately it seems that Performer inst-alls pieces of itself in a large number of places, so that one would have to mount a large number of rather high-level trees to all the machines, clearly an undesirable situation. Is there a way to install Performer in a single dir tree (/usr/Performer say) and export/mount that and maybe set a few env variables.... that sort of thing? It would both save a stack of disk space and make installation/upgrades easier (would only have to be done on the server). ANy ideas? Carlo Tiana NASA Ames Research Center From guest Thu Aug 11 13:43:22 1994 Date: Thu, 11 Aug 94 12:39:54 -0500 From: petert@killerbee.jsc.nasa.gov (Peter Tran) To: info-performer@sgi.sgi.com Subject: Can Performer runs on Indy? Hi everyone, I've two questions that I want to ask about Performer. 1) What is the minimum HW/SW requirements to run Performer? 2) Can you run Performer on the Indy? If so, what are the minimum requirements? I'd appreciate all the comments I can get from your helps. Thanks in advance, P.T From guest Thu Aug 11 13:50:05 1994 Date: Wed, 10 Aug 1994 17:34:42 -0700 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: info-performer@sgi.sgi.com From: jan@archimedes.chinalake.navy.mil (Jan A. Barglowski) Subject: Ptu (Performer Terrain Utility) samples? Does anyone use the PTU loader? If so, can I possible get a .ptu file, and the corresponding terrain image and overlays? The manual says the terrain height is calculated by the brightness of the monochromatic image. Is this a limitation to between 0-255, or can you use a 16-bit mono image to get better resolution? In looking at the ptu loader code and the description of parameters in the manual, they don't seem to be the same. Something for the doc department... For government users, I just got a copy of a program called MUSE, from the DMA. It allows various sort of maps (imagery) to be created from just about any DMA data format. I was able to extract out a nice tiff image of an exact area from an ADRG CDROM in just a few minutes... jan. Jan Anthony Barglowski Phone: (619) 927-1057 Computer Dude Internet: jan@archimedes.nwc.navy.mil Visualization Lab Packet: SYSOP@WA6YBN.#SOCA.CA.US.NA Naval Air Warfare Center kc6uth@kc6uth.ampr.org -Weapons Division [44.17.2.6] From guest Thu Aug 11 13:47:26 1994 To: "Carlo L. Tiana" Cc: info-performer@sgi.sgi.com, seguin@vr1.engin.umich.edu Subject: Re: network installation In-Reply-To: Your message of "Wed, 10 Aug 1994 21:26:32 GMT." <199408102126.VAA08168@fechner.arc.nasa.gov> Date: Thu, 11 Aug 1994 10:22:25 -0400 From: Ralph Seguin > We run performer on a number of machines (Onyx and Indigo 2) and would > like to avoid having a full distribution of Performer on each machine, > but rather would like to mount it from the server for all the machines. > Unfortunately it seems that Performer inst-alls pieces of itself in a > large number of places, so that one would have to mount a large number > of rather high-level trees to all the machines, clearly an undesirable > situation. > Is there a way to install Performer in a single dir tree (/usr/Performer > say) and export/mount that and maybe set a few env variables.... that > sort of thing? > It would both save a stack of disk space and make installation/upgrades > easier (would only have to be done on the server). I asked this same question a while back (about multiple packages though). There is one "gross" way of doing it. Install it on every machine, but have the installation tree be a shared location. Eg: machines: m1, m2, m3, m4 m1 has /disk1 exported NFS to m2, m3, m4 m2, m3, m4 automount /disk1 as /nfs/m1/disk1 on each of m2, m3, m4, make a symlink from /nfs/m1/disk1/swroot into /usr/swroot on m1, make a symlink from /disk1/swroot into /usr/swroot now, on every machine, run inst like so... inst -f /CDROM/dist -r /usr/swroot advantages: one copy shared disadvantages: takes LOTS of time, and lots of repeated work I agree that there should be an easier way of doing this, but SGI told me that there wasn't. Ideally, there would be something that would run strictly the finish up routines from inst (ie, to generate the necessary symlinks, ...) -Ralph From guest Thu Aug 11 13:47:53 1994 (5.65c/IDA-1.4.4 for <@datasrv.co.il:info-performer@sgi.com>); Thu, 11 Aug 1994 08:54:15 +0300 From: "Ran Yakir" Date: Thu, 11 Aug 1994 08:51:44 +0000 In-Reply-To: "Ekkehard Klaerner" "pfTranslate" (Aug 10, 5:20pm) References: <71E0A743789@aitec.de> X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: "Ekkehard Klaerner" , info-performer@sgi.sgi.com Subject: Re: pfTranslate Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 On Aug 10, 5:20pm, Ekkehard Klaerner wrote: > Subject: pfTranslate > > Hi , > > Is the use of pfTranslate and pfRotate in nodeDrawCallbacks faster > than pfSCS, if we just need translation and rotation ? Note that the Draw process normally processes a display list, prepared in the Cull process. I suppose that the display list contains the pushmatrix(), translate(), rotate(), etc for the SCS already. If you insert draw callbacks, they are bound to be called from inside the processing of the display list. I don't think there is a chance that this thing will not hurt you. Function call takes time too. From my experience, node callbacks create a lot of overhead if they occur many times in a frame. Ran -- __ | Ran Yakir /_) _ __ \ / _ / o __ | Graphics App. Chief Engineer / )_ (_(_) ) \/ (_(_/<_(_)( | BVR Technologies Ltd. _/ | -------------------------------------+-------------------------------- Phone : | E-mail : rany@bvr.co.il Work : 972-3-5715671 | Res. : 972-3-6995364 | Fax : 972-3-5715668 | From guest Thu Aug 11 13:46:38 1994 Date: Thu, 11 Aug 94 09:19:56 PDT From: goodhart@cod.nosc.mil (Curtis L. Goodhart) To: info-performer@sgi.sgi.com Subject: MultiProcessing and heavy computation Cc: goodhart@cod.nosc.mil, ling@cod.nosc.mil Currently my application spends a lot of time in the draw process. We have considered a solution to a problem that would involve computing the distance to every triangle in the scene. This seems crazy at first because of the amount of computation involved. However my application spends much more time in the draw process than anywhere else, perhaps as much as 10x or even more. Is there any hope that the distances computations could be done in a reasonable time in a separate application or even cull process, taking advantage of the parallel processing while the lengthy draw goes on anyway, or is the graphics hardware so fast that there would be no hope of keeping up? As far as sizing things I think the main point is that we want to compute the distance to as many polygons as will be drawn (I think we draw anywhere from hundreds to thousands of polygons and have some texture mapping too. We are running on a 4 processor ONYX. Thanks, Curt Goodhart From guest Thu Aug 11 13:46:43 1994 Date: Thu, 11 Aug 1994 09:07:04 -0700 From: hitchner@netcom.com (Lew Hitchner) To: carlo@vision.arc.nasa.gov, info-performer@sgi.sgi.com Subject: Re: network installation Carlo Tiana of NASA Ames writes: >> We run performer on a number of machines (Onyx and Indigo 2) and would >> like to avoid having a full distribution of Performer on each machine, >> but rather would like to mount it from the server for all the machines. . . . >> Is there a way to install Performer in a single dir tree (/usr/Performer >> say) and export/mount that and maybe set a few env variables.... that >> sort of thing? How about this? 1. create a file system, say /usr/performer, on the host server -- note: don't name it /usr/Performer 2. NFS mount that file system to all other machines 3. Create soft links between the directories on all the machines, including the server, to the separate Performer files and directories, e.g., % ln -s /usr/performer/lib/libp*.a /usr/lib % ln -s /usr/performer/src /usr/Performer/src I haven't tried this, tho we've done similar things with other file systems. Potential problems: - I'm not sure how this will interact with installing Performer. Obvioulsy, the links must be created before installing. So, that might mean if you've already installed Performer, you might have to uninstall it, create the links and new fs, then reinstall. - This probably isn't possible if if your machines aren't running the same version of IRIX. Lew Hitchner Virtual Reality and Visual Simulation Consultant Mountain View, CA From guest Thu Aug 11 14:02:19 1994 Date: Thu, 11 Aug 94 09:09:23 PDT From: goodhart@cod.nosc.mil (Curtis L. Goodhart) To: info-performer@sgi.sgi.com Subject: MultiProcessing and Intersections Cc: goodhart@cod.nosc.mil, ling@cod.nosc.mil On page of 177 of the User's Guide Table 7-4 syas that pfSegsIsectNode should only be called from an ISECT or APP process. I dynamically determine an object to highlight using pfChanPick (which I believe calls pfSegsIsectNode). I do this from from a cull callback function. I guess that's why it chokes and dies (from a child dying) ie because I am calling intersections in a CULL process and I am multiprocessing. When I put the pfChanPick in the application section it doesn't die but the highlighting gets flaky and seems to jump from object to object rather than stay on the object that I expect to be highlighted. Sometimes objects that are ont supposed to be selectable for highlighting get highlighted. 1) When and where does the ISECT process occur especially in terms of multiprocessing and the APP/CULL/DRAW model. 2) Any ideas on why the dynamically selected object seems to get flaky? 3) Does putting this call in the APP process seem the correct thing to do? Are there other choices that would make more sense? We don't want it in the draw process because it is the big time user already. Thanks, Curt Goodhart From guest Thu Aug 11 15:15:49 1994 From: "Michael Jones" Date: Thu, 11 Aug 1994 15:15:25 -0700 In-Reply-To: petert@killerbee.jsc.nasa.gov (Peter Tran) "Can Performer runs on Indy?" (Aug 11, 12:39pm) References: <9408111739.AA18223@killerbee.jsc.nasa.gov> X-Mailer: Z-Mail (3.2a.627 27jun94 MediaMail) To: petert@killerbee.jsc.nasa.gov (Peter Tran), info-performer@sgi.sgi.com Subject: Re: Can Performer runs on Indy? Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii On Aug 11, 12:39pm, Peter Tran wrote: > Subject: Can Performer runs on Indy? : 1) What is the minimum HW/SW requirements to run Performer? Indy with 5.2 or Personal Iris with 4.0.5 : 2) Can you run Performer on the Indy? If so, what are the minimum : requirements? Yes. It will run on any Indy. Graphics on basic Indy machines are not very fast compared to the VGXT and RealityEngine, though, so don't expect 30 Hz on the Performer Town. We've not yet seen one of the hardware accelerated Indy's so I can't comment on the speeds you would get on one of those. -- Be seeing you, Phone:415.390.1455 Fax:415.390.2658 M/S:8U-590 Michael T. Jones Silicon Graphics, Advanced Graphics Division mtj@sgi.com 2011 N. Shoreline Blvd., Mtn. View, CA 94039-7311 From guest Thu Aug 11 17:40:01 1994 From: "Chris Tanner" Date: Thu, 11 Aug 1994 17:39:24 -0700 In-Reply-To: jan@archimedes.chinalake.navy.mil (Jan A. Barglowski) "Ptu (Performer Terrain Utility) samples?" (Aug 10, 5:34pm) References: <199408110034.RAA01381@archimedes.chinalake.navy.mil> X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: jan@archimedes.chinalake.navy.mil (Jan A. Barglowski), info-performer@sgi.sgi.com Subject: Re: Ptu (Performer Terrain Utility) samples? Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 On Aug 10, 5:34pm, Jan A. Barglowski wrote: > Subject: Ptu (Performer Terrain Utility) samples? > Does anyone use the PTU loader? If so, can I possible get a .ptu file, and > the corresponding terrain image and overlays? > > The manual says the terrain height is calculated by the brightness of the > monochromatic image. Is this a limitation to between 0-255, or can you use > a 16-bit mono image to get better resolution? > > In looking at the ptu loader code and the description of parameters in the > manual, they don't seem to be the same. Something for the doc > department... > > For government users, I just got a copy of a program called MUSE, from the > DMA. It allows various sort of maps (imagery) to be created from just > about any DMA data format. I was able to extract out a nice tiff image of > an exact area from an ADRG CDROM in just a few minutes... > > jan. > > > Jan Anthony Barglowski Phone: (619) 927-1057 > Computer Dude Internet: jan@archimedes.nwc.navy.mil > Visualization Lab Packet: SYSOP@WA6YBN.#SOCA.CA.US.NA > Naval Air Warfare Center kc6uth@kc6uth.ampr.org > -Weapons Division [44.17.2.6] > > > > >-- End of excerpt from Jan A. Barglowski The format in the programming guide is missing a field... The real format is... TerrainNodeName #LODs /* Typically 3-5 */ #TilesX #TilesY /* The more the better (4-10,4-10) */ ScaleX ScaleY ScaleZ /* Pixel->Performer Unit description of elevation data */ Elevation.bw SizeX SizeY /* bw or fit data 8 or 16 bit *should* work*/ /* Note sizeX sizeY tell Performer the highest resolution */ /* Of the data across the whole terrain - Performer will use */ /* ImageVision library to downsample to this resolution and */ /* tile the data */ Satellite.rgb SizeX SizeY /* Specifies an RGB image to drape over the elevation data */ /* Once again SizeX SizeY specify the resolution the image */ /* should be sampled down to before tiling occurs */ /* IL is used for this as well */ DetailTexture.rgb /* Optional detail texture and detail texture spline */ /* Have seen some anomolies when using detail texture...*/ A Sample File ("ystone.ptu") yellowstone 4 8 8 200.0 200.0 5.0 /usr/demos/data/Performer/terrain/ystone1024.bw 1024 1024 /usr/demos/data/Performer/terrain/ystone2048.rgb 2048 2048 detail.rgb where... cct@faith>> istat /hosts/rose/d2/people/demos/data/Performer/terrain/ystone2048.rgb xsize ysize zsize min max bpp type storage name 2048 2048 3 0 255 1 NORMAL verb ystone2048.rgb cct@faith>> istat /hosts/rose/d2/people/demos/data/Performer/terrain/ystone1024.bw xsize ysize zsize min max bpp type storage name 1024 1024 1 0 3712 2 NORMAL verb ystone1024.bw The ptu stuff isnt perfect, but it makes it easy to fly over terrain data getting decent performance without much effort. (Remember that its just utility stuff though and isnt intend to be THE solution, rather its just a helpful utility) Later, Chris Tanner IRIS Performer -- _____________________________________________________________ Chris Tanner (cct@faith.asd.sgi.com) Silicon Graphics - Advanced Graphics Division _____________________________________________________________ From guest Fri Aug 12 05:09:11 1994 From: wags@math.tau.ac.il Date: Fri, 12 Aug 94 15:07:46 +0300 To: info-performer@sgi.sgi.com Subject: Unsubscribe I know you're not supposed to post this in the mailing list but I don't know the username to which yor are supposed to send this to. Thanks & sorry. wags. From guest Fri Aug 12 08:46:04 1994 From: hodeckes@VNET.IBM.COM Fri, 12 Aug 94 11:46:26 EDT Date: Fri, 12 Aug 94 11:45:34 EDT To: info-performer@sgi.sgi.com Subject: Modifying geosets Hi all, First, has anyone used the ImageVision Library with Performer. Please comment on any issues that must be addressed to do this. Secondly, I am having difficulty determining how to modify/edit the structure of a geoset. Currently I must regenerate the entire geoset to make a change to the underlying geometry. For example, I create a geoset of linestrips between a given set of points. If the coordinates of one of the points change, I want to modify that single vertex of the linestrip instead of regenerating the entire linestrip. I guess what I am looking for is a handle to the geoset structure. Thanks for your help. Steve Hodecker LORAL Federal Systems Flight Systems Engineering From guest Fri Aug 12 09:50:37 1994 From: halliday@BanffCentre.AB.CA (Sean Halliday) Subject: Re: Modifying geosets To: hodeckes@vnet.ibm.com Date: Fri, 12 Aug 1994 10:49:41 -0700 (MDT) Cc: info-performer@sgi.sgi.com In-Reply-To: <199408121546.IAA05140@sgi.sgi.com> from "hodeckes@vnet.ibm.com" at Aug 12, 94 11:45:34 am X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1095 > > Hi all, > > First, has anyone used the ImageVision Library with Performer. Please > comment on any issues that must be addressed to do this. > > Secondly, I am having difficulty determining how to modify/edit the > structure of a geoset. Currently I must regenerate the entire geoset to > make a change to the underlying geometry. > > For example, I create a geoset of linestrips between a given set of > points. If the coordinates of one of the points change, I want to > modify that single vertex of the linestrip instead of regenerating the > entire linestrip. > > I guess what I am looking for is a handle to the geoset structure. > Thanks for your help. > > Steve Hodecker > LORAL Federal Systems > Flight Systems Engineering > > > You must pfMalloc the vertices before pfConfig. This allows you to modify the vertix array. (see Shared memory debugging in the manual) If you pfMalloc afterwords, each process will get a different pointer. -- Sean Halliday Computer Graphics Software Developer New Media Research, Banff Centre for the Arts. halliday@BanffCentre.AB.CA From guest Fri Aug 12 10:19:35 1994 Date: Fri, 12 Aug 1994 13:20:21 -0400 (EDT) From: Kevin Mueller Subject: Cull'ing To: info-performer Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi there: I'm having some CULL problems. I have created a simple terrain using triangle mesh. The terrain is tiled with each tile belonging to a different geometry node. As the view is traveling across the terrain, tiles in the viewing frustum will be culled and not displayed. I've added post cull callbacks to the pfGeode. In the callback, pfGetCullResult() returns zero, pfGetNodeTravData() returns all 1's for CULL and DRAW mask, and the sphere returned by pfGetNodeBSphere() is set up properly. I can get the tile's to "pop" in and out just by modifying the view heading. I've run out of ideals to debug what is going on. Any help would be appreciated. -- Kevin Mueller email: mueller@vsl.ist.ucf.edu Visual Systems Scientist desk: 407-249-4711 University of Central Florida lab: 407-249-4703/4704 Institute for Simulation and Training fax: 407-658-5059 3280 Progress Drive Orlando, FL 32826 From guest Fri Aug 12 12:34:10 1994 From: fraser@portola.com (Glen D. Fraser) id AA10589 for info-performer@sgi.com; Fri, 12 Aug 94 12:34:18 PDT Subject: Re: Modifying geosets To: hodeckes@VNET.IBM.COM Date: Fri, 12 Aug 1994 12:34:17 -0700 (PDT) Cc: info-performer@sgi.sgi.com In-Reply-To: <199408121546.IAA05140@sgi.sgi.com> from "hodeckes@VNET.IBM.COM" at Aug 12, 94 11:45:34 am X-Mailer: ELM [version 2.4 PL21] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 845 > Hi all, > > First, has anyone used the ImageVision Library with Performer. Please > comment on any issues that must be addressed to do this. [ stuff about geoset modification deleted] > Steve Hodecker > LORAL Federal Systems > Flight Systems Engineering You might want to look at the code for the Performer Terrain Utilities (PTU). This database loader makes use of the ImageVision Library. Check out /usr/src/Performer/src/lib/libpfsgi/pfptu.c. =========================================================== Glen D. Fraser Head, Software and Telepresence Research, Inc. System Development 320 Gabarda Way Phone: (415)854-4420 Portola Valley, CA 94028 Fax: (415)854-3141 USA E-Mail: fraser@telepresence.com =========================================================== From guest Fri Aug 12 12:54:11 1994 To: halliday@banffcentre.ab.ca (Sean Halliday) Cc: hodeckes@vnet.ibm.com, info-performer@sgi.sgi.com Subject: Re: Modifying geosets In-Reply-To: Your message of "Fri, 12 Aug 94 10:49:41 PDT." <9408121649.AA18247@grizzly.BanffCentre.AB.CA> Date: Fri, 12 Aug 94 12:53:03 -0700 From: Jim Helman Vertex, normal, color and texture coordinate arrays are fully exposed in the array you (or the loader) provided to pfGSetAttrLists, so if you want to change a vertex, just change it. There's no rendering penalty for the change, since Performer usually renders in immediate mode straight out of these arrays. If the change affects the bounding box, you may want to do pfGSetBBox(gset, NULL) to reset it or use a large static bounding box. If you need to change the topology or number of vertices, you should allocate the largest size necessary up front to avoid run-time mallocs, if possible. In an MP application, you may wish to multibuffer the geometry using multiple copies under a pfswitch node to avoid modifying something in the APP process while the DRAW process is rendering it. The vertex arrays can be pfMalloced at any time. Sean's point concerns passing array pointers through global variables not in shared memory, but one can easily use shared memory to pass pointers of arrays allocated after pfConfig. rgds, -jim helman jimh@surreal.asd.sgi.com 415/390-1151 From guest Fri Aug 12 13:33:40 1994 ; Fri, 12 Aug 1994 16:32:54 -0400 Date: Fri, 12 Aug 94 13:31:25 -0700 From: jarch@gemtech.com (John L. Archdeacon) To: info-performer@sgi.sgi.com Subject: Unsubscribe me please Please unsubscribe me! Thanks. -- jla@gemtech.com -- (John L. Archdeacon, Gemini Technology Corporation ; 714-727-1980) From guest Fri Aug 12 13:33:46 1994 ; Fri, 12 Aug 1994 16:32:56 -0400 Date: Fri, 12 Aug 94 13:32:21 -0700 From: pfnews@gemtech.com (IRIS Performer News) To: info-performer@sgi.sgi.com Subject: Please subscribe me Please subscribe me! Thanks. -- pfnews@gemtech.com -- (Gemini Technology Corporation ; 714-727-1980) From guest Mon Aug 15 14:43:26 1994 Date: Mon, 15 Aug 94 23:43 MDT From: matomira@lig.di.epfl.ch (Fernando D. Mato Mira) To: info-performer@sgi.sgi.com Subject: projection matrices Reply-To: matomira@epfl.ch Hello, Arent' there any routines to get and set pfFrustum projection matrices directly? Thanks, Fernando D. Mato Mira Computer Graphics Lab Swiss Federal Institute of Technology (EPFL) Phone : +41 (21) 693 - 5248 CH-1015 Lausanne FAX : +41 (21) 693 - 5328 Switzerland E-mail : matomira@di.epfl.ch From guest Wed Aug 17 05:34:32 1994 (5.65c/IDA-1.4.4 for <@datasrv.co.il:info-performer@sgi.com>); Wed, 17 Aug 1994 15:35:17 +0300 From: "Aviva Chavel" Date: Wed, 17 Aug 1994 15:32:55 +0000 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: info-performer@sgi.sgi.com Subject: multigen/modelgen Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 I would like to know what differences, if any, exist between Multigen and ModelGen. For instance, are there features supported by Multigen that are unsupported in ModelGen, and vice versa. Thank you. Aviva Chavel From guest Wed Aug 17 06:07:20 1994 (5.65c/IDA-1.4.4 for ); Wed, 17 Aug 1994 09:07:13 -0400 Date: Wed, 17 Aug 94 09:06:42 EDT From: fyock@capella.tsc.gtefsd.com (Christina Fyock) To: aviva@bvr.co.il, info-performer@sgi.sgi.com Subject: Re: multigen/modelgen Modelgen is a scaled down version of Multigen, hence it is less expensive. The company has just announced a new product, called Modelgen2 that offers more options such as: Terrain, Instrumentation, Behavior and 3D Audio. The original Modelgen just had a texture and vr option. The number for Multigen Inc is 408-247-4326, you can call them for more specific information and pricing. From guest Wed Aug 17 06:03:47 1994 (5.65c/IDA-1.4.4 for <@datasrv.co.il:info-performer@sgi.com>); Wed, 17 Aug 1994 16:03:55 +0300 From: "Ran Yakir" Date: Wed, 17 Aug 1994 16:01:27 +0000 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: info-performer@sgi.sgi.com Subject: Bug in pfuTraverse Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Hi All I don't remember if I wrote about that bug before. Anyway, there is a bug in the PFUTRAV_LOD_RANGE0 mode of pfuTraverse. When called using taht mode, pfuTraverse should enter only the best LOD in an LOD node. However, it enters no LOD at all. The reason for that is propobaly a typo in the trav.c code : case PFTYPE_LOD: switch (trav->mode & PFUTRAV_LOD_MASK) { case PFUTRAV_LOD_RANGE0: range0 = pfGetLODRange((pfLOD *)node, 0); range1 = pfGetLODRange((pfLOD *)node, 0); if (range0 <= 0.0f && 0.0f < range1) if (numChild > 0) ret = pfuTraverse(pfGetChild(node, 0), trav); break; ... ... } Ofcourse, in that case, range0 and range1 are equal (both taken from LODrange 0). The condition can never be true, and no traversal is done for this node. To fix replace range1 = pfGetLODRange((pfLOD *)node, 0); with range1 = pfGetLODRange((pfLOD *)node, 1); Regards Ran Yakir -- __ | Ran Yakir /_) _ __ \ / _ / o __ | Graphics App. Chief Engineer / )_ (_(_) ) \/ (_(_/<_(_)( | BVR Technologies Ltd. _/ | -------------------------------------+-------------------------------- Phone : | E-mail : rany@bvr.co.il Work : 972-3-5715671 | Res. : 972-3-6995364 | Fax : 972-3-5715668 | From guest Wed Aug 17 13:20:27 1994 ; Wed, 17 Aug 1994 16:20:13 -0400 id AA01788; Wed, 17 Aug 1994 13:24:45 PST Organization: MultiGen Inc X-Charset: MACINTOSH X-Umcp-To: INFO PERFORMER From: Marcus To: INFO PERFORMER Date: Wed, 17 Aug 1994 12:14:53 PST Subject: 2-color directional lightpo REGARDING 2-color directional lightpoints Hi All, Let's say we have a string of bi-directional light points with 4 points. And they all have the same direction and a front color of red. All but one have a back color of red as well. The odd ball light has a back color of white. Like so: FRONT | red | red | red | red ------------------------------- BACK | red | red | red | white Performer cannot have a bidirectional light with different colors. So we want to make opposing pairs of unidirectional lights to simulated this, where neccesary. The performance question is: Would it be better to make 1 bidirectional pfLightPoint for the 3 red lights, plus 2 unidirectional pfLightPoints (1 red and 1 white)? OR Would it be better to make 1 unidirectional pfLightPoint for the 4 red lights, plus 1 unidirectional pfLightPoint for the 3 red and 1 white light? Basically, which is more efficient: 3 pfLightPoints of 5 lights, or 2 pfLightPoints of 8 lights? Also ... consider the trade-off on a much larger scale. Right now I'm leaning towards more lights per pfLightPoint. Thanks in advance, Marcus Barnes, Member Technical Staff MultiGen Inc., 1884 The Alameda, San Jose CA, 95126 PH: (408) 261 4110 FX: (408) 247 4329 EMAIL: multigen!marcus@uunet.UU.NET From guest Wed Aug 17 14:17:16 1994 Date: Wed, 17 Aug 94 23:17 MDT From: matomira@lig.di.epfl.ch (Fernando D. Mato Mira) To: info-performer@sgi.sgi.com Subject: texgen Reply-To: matomira@epfl.ch Hello, The man page for pfTexture says: >> For geometry to be textured, the following must be true: 1. Texturing must be enabled: pfEnable(PFEN_TEXTURE) 2. A pfTexEnv must be applied: pfApplyTEnv 3. A pfTexture must be applied: pfApplyTex 4. Geometry must have texture coordinates: pfGSetAttr, PFGS_TEXCOORD2 >> Doesn't texgen work also for point 4. ? And then, I guess TG_ON and TG_OFF is my own responsibility (so that I can override the coords in the pfGeoSet).. Thanks, Fernando D. Mato Mira Computer Graphics Lab Swiss Federal Institute of Technology (EPFL) Phone : +41 (21) 693 - 5248 CH-1015 Lausanne FAX : +41 (21) 693 - 5328 Switzerland E-mail : matomira@di.epfl.ch From guest Wed Aug 17 14:18:52 1994 Date: Wed, 17 Aug 94 07:18:28 EST From: elkins@degas.nswc.navy.mil (Les Elkins b40 x43850) To: info-performer@sgi.sgi.com Subject: ImageVision w/ptu The discussion of the ptu loader on this list last week aroused my interest. However, I'm not very knowledgeable about the ImageVision Library product. I'm running under 5.2, and I do have the il_eoe and imgtools subsystems installed. I have been trying to recompile the libpfsgi code but I seem to be missing some .h files for ImageVision. Are these lurking on one of the other 5.2 distribution cd's, or do I need to buy some kind of ImageVision developement system? Any info would be helpful... Les Elkins ---------------------------------------------------------------------- lelkins@relay.nswc.navy.mil The views expressed herein do not represent those of NSWC, the Navy, Les Elkins or the federal government. Naval Surface Warfare Center Dahlgren Division (And anybody who says otherwise is Silver Spring, MD itching for a fight...) From guest Wed Aug 17 16:08:06 1994 From: watsen@netcom.com (Kent Watsen) Subject: Cloning pfGSetAttr To: info-performer@sgi.sgi.com (performer) Date: Wed, 17 Aug 1994 16:08:25 -0700 (PDT) X-Mailer: ELM [version 2.4 PL23] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 447 Hello, Given a pointer to a pfNode which is the root of some geometry, will a pfClone of this node allow me to change the colors of some polygons without affecting the the colors of the original or any other pfClones of the original? If Not, if there another way to copy all of the geometry (LOD, DCS, pfGSetAttr, etc) of a given node without having to reload it from disk? Kent Watsen DCS Corporation Simulation Branch 703.683.8430 x369 From guest Wed Aug 17 16:20:01 1994 From: "Chris Tanner" Date: Wed, 17 Aug 1994 16:19:43 -0700 In-Reply-To: matomira@lig.di.epfl.ch (Fernando D. Mato Mira) "texgen" (Aug 17, 11:17pm) References: X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: matomira@epfl.ch, info-performer@sgi.sgi.com Subject: Re: texgen Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 On Aug 17, 11:17pm, Fernando D. Mato Mira wrote: > Subject: texgen > > Hello, > > The man page for pfTexture says: > > >> > For geometry to be textured, the following must be true: > > 1. Texturing must be enabled: pfEnable(PFEN_TEXTURE) > > 2. A pfTexEnv must be applied: pfApplyTEnv > > 3. A pfTexture must be applied: pfApplyTex > > 4. Geometry must have texture coordinates: pfGSetAttr, > PFGS_TEXCOORD2 > >> > > Doesn't texgen work also for point 4. ? > And then, I guess TG_ON and TG_OFF is my own responsibility (so > that I can override the coords in the pfGeoSet).. > > Thanks, > > Fernando D. Mato Mira > Computer Graphics Lab > Swiss Federal Institute of Technology (EPFL) Phone : +41 (21) 693 - 5248 > CH-1015 Lausanne FAX : +41 (21) 693 - 5328 > Switzerland E-mail : matomira@di.epfl.ch > > >-- End of excerpt from Fernando D. Mato Mira Its very easy to use texgen with Performer. You simply insert callbacks on nodes that you want to be textured via texgen. That way since you specify that these callbacks should happen in the "DRAW" traversal, you can turn on and off texgen as appropriate before and after drawing the geometry in question. For instance if you had a node named "geodeRefl" on which you wanted to use texgen to create pseudo environment mapping, then you would make the following call on the node in order to setup a "draw callback" where "pfuPreDrawReflMap" is a function that turns texgen for sphere maps on and "pfuPostDrawReflMap" turns it off. pfNodeTravFuncs(geodeRefl, PFTRAV_DRAW, pfuPreDrawReflMap, pfuPostDrawReflMap); These functions are even already defined in the libpfutil library in "callbacks.c" in /usr/src/Performer/src/lib/libpfutil/. However, if you would like to use other types of texgen (ie linear or contour mapping), then replace that file with this one that defines Pre and Post draw callbacks for Linear,Contour, and Environment mapping... callbacks.c: /* * Copyright (c) 1994 Silicon Graphics, Inc. * * Permission to use, copy, modify, distribute, and sell this software * and its documentation for any purpose is hereby granted without * fee, provided that (i) the above copyright notices and this * permission notice appear in all copies of the software and related * documentation, and (ii) the name of Si MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. * * IN NO EVENT SHALL SILICON GRAPHICS BE LIABLE FOR ANY SPECIAL, * INCIDENTAL, INDIRECT OR CONSEQUENTIAL DAMAGES OF ANY KIND, OR ANY * DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, * WHETHER OR NOT ADVISED OF THE POSSIBILITY OF DAMAGE, AND ON ANY * THEORY OF LIABILITY, ARISING OUT OF OR IN CONNECTION WITH THE USE * OR PERFORMANCE OF THIS SOFTWARE. * */ #include #include #include #include "pf.h" #include "pfutil.h" #include "image.h" static char CallBacks[] = "CallbacksDPool"; static pfMatrix IdentMat = { {1.0f,0.0f,0.0f,0.0f}, {0.0f,1.0f,0.0f,0.0f}, {0.0f,0.0f,1.0f,0.0f}, {0.0f,0.0f,0.0f,1.0f} }; void pfuPrintTexgenParams(void) { pfDataPool *dp; float *paramsX, *paramsY; dp = pfAttachDPool(CallBacks); if (dp == NULL) pfuTexgenParams(NULL, NULL); paramsX = (float *)pfDPoolFind(dp, 1); paramsY = (float *)pfDPoolFind(dp, 2); fprintf(stderr,"ParamsX: %f\t%f\t%f\t%f\n", paramsX[0],paramsX[1],paramsX[2],paramsX[3]); fprintf(stderr,"ParamsY: %f\t%f\t%f\t%f\n", paramsY[0],paramsY[1],paramsY[2],paramsY[3]); } void pfuGetTexgenParams(float *newParamsX, float *newParamsY) { pfDataPool *dp; float *paramsX, *paramsY; int i; dp = pfAttachDPool(CallBacks); if (dp == NULL) pfuTexgenParams(NULL, NULL); paramsX = (float *)pfDPoolFind(dp, 1); paramsY = (float *)pfDPoolFind(dp, 2); for (i=0;i<4;i++) newParamsX[i] = paramsX[i]; for (i=0;i<4;i++) newParamsY[i] = paramsY[i]; } void pfuTexgenParams(float *newParamsX, float *newParamsY) { pfDataPool *dp; float *paramsX, *paramsY; dp = pfAttachDPool(CallBacks); if (dp == NULL) { dp = pfNewDPool(sizeof(float) * 8, CallBacks); paramsX = (float *)pfDPoolAlloc(dp, sizeof(float) * 4, 1); paramsY = (float *)pfDPoolAlloc(dp, sizeof(float) * 4, 2); } else { paramsX = (float *)pfDPoolFind(dp, 1); paramsY = (float *)pfDPoolFind(dp, 2); } if (!newParamsX) { paramsX[0] = 0.0; paramsX[1] = 1.0; paramsX[2] = 0.0; paramsX[3] = 1.0; } else { paramsX[0] = newParamsX[0]; paramsX[1] = newParamsX[1]; paramsX[2] = newParamsX[2]; paramsX[3] = newParamsX[3]; } if (!newParamsY) { paramsY[0] = 0.0; paramsY[1] = 1.0; paramsY[2] = 0.0; paramsY[3] = 1.0; } else { paramsY[0] = newParamsX[0]; paramsY[1] = newParamsX[1]; paramsY[2] = newParamsX[2]; paramsY[3] = newParamsX[3]; } } long pfuPreDrawContourMap(pfTraverser *trav, void *data) { pfDataPool *dp; float *paramsX, *paramsY; dp = pfAttachDPool(CallBacks); if (dp == NULL) pfuTexgenParams(NULL, NULL); paramsX = (float *)pfDPoolFind(dp, 1); paramsY = (float *)pfDPoolFind(dp, 2); (trav, trav); (data, data); mmode(MVIEWING); pushmatrix(); loadmatrix(IdentMat); texgen(TX_S, TG_CONTOUR, paramsX); texgen(TX_T, TG_CONTOUR, paramsY); texgen(TX_S, TG_ON, NULL); texgen(TX_T, TG_ON, NULL); popmatrix(); return NULL; } long pfuPostDrawContourMap(pfTraverser *trav, void *data) { (trav, trav); (data, data); texgen(TX_S, TG_OFF, NULL); texgen(TX_T, TG_OFF, NULL); return NULL; } long pfuPreDrawLinearMap(pfTraverser *trav, void *data) { pfDataPool *dp; float *paramsX, *paramsY; dp = pfAttachDPool(CallBacks); if (dp == NULL) pfuTexgenParams(NULL, NULL); paramsX = (float *)pfDPoolFind(dp, 1); paramsY = (float *)pfDPoolFind(dp, 2); (trav, trav); (data, data); texgen(TX_S, TG_SPHEREMAP, paramsX); texgen(TX_T, TG_SPHEREMAP, paramsY); texgen(TX_S, TG_ON, NULL); texgen(TX_T, TG_ON, NULL); return NULL; } long pfuPostDrawLinearMap(pfTraverser *trav, void *data) { (trav, trav); (data, data); texgen(TX_S, TG_OFF, NULL); texgen(TX_T, TG_OFF, NULL); return NULL; } long pfuPreDrawReflMap(pfTraverser *trav, void *data) { (trav, trav); (data, data); texgen(TX_S, TG_SPHEREMAP, 0); texgen(TX_T, TG_SPHEREMAP, 0); texgen(TX_S, TG_ON, NULL); texgen(TX_T, TG_ON, NULL); return NULL; } long pfuPostDrawReflMap(pfTraverser *trav, void *data) { (trav, trav); (data, data); texgen(TX_S, TG_OFF, NULL); texgen(TX_T, TG_OFF, NULL); return NULL; } -- _____________________________________________________________ Chris Tanner (cct@faith.asd.sgi.com) Silicon Graphics - Advanced Graphics Division _____________________________________________________________ From guest Wed Aug 17 16:40:20 1994 From: "Chris Tanner" Date: Wed, 17 Aug 1994 16:40:02 -0700 In-Reply-To: elkins@degas.nswc.navy.mil (Les Elkins b40 x43850) "ImageVision w/ptu" (Aug 17, 7:18am) References: <9408171218.AA01017@degas.nswc.navy.mil> X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: elkins@degas.nswc.navy.mil (Les Elkins b40 x43850), info-performer@sgi.sgi.com Subject: Re: ImageVision w/ptu Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 On Aug 17, 7:18am, Les Elkins b40 x43850 wrote: > Subject: ImageVision w/ptu > > The discussion of the ptu loader on this list last week aroused my > interest. However, I'm not very knowledgeable about the ImageVision > Library product. I'm running under 5.2, and I do have the il_eoe and > imgtools subsystems installed. I have been trying to recompile the > libpfsgi code but I seem to be missing some .h files for ImageVision. > Are these lurking on one of the other 5.2 distribution cd's, or do > I need to buy some kind of ImageVision developement system? > > Any info would be helpful... > > Les Elkins > > ---------------------------------------------------------------------- > lelkins@relay.nswc.navy.mil The views expressed herein do not > represent those of NSWC, the Navy, > Les Elkins or the federal government. > Naval Surface Warfare Center > Dahlgren Division (And anybody who says otherwise is > Silver Spring, MD itching for a fight...) > > >-- End of excerpt from Les Elkins b40 x43850 Because you are actually compiling IL code, you do need to purchase the development option for IL. IL is very useful and if you are going to lots of image processing / image manipulation its definitely worth purchasing. However, I just placed the IL compiled version of the libpfsgi library on sgi.com in ~ftp/pub/libpfsgi{-g}.a (this should enable anonymous ftp)... If you compile perfly with one of these library's, then the il_eoe subsytem of IL should be enough to run the ptu loader. If you want to change/fix anything in the ptu loader you should get the il_dev option. Ahha! I put ilgettile.o and ilgettile-g.o in the same directory on sgi.com. This should enable you to fix/modify everything about ptu except the actual way it retrieves tiles using IL. (just use ilgettile.o and never modify ilgettile.c) Hope that helps everyone... Later, Chris Tanner IRIS Performer -- _____________________________________________________________ Chris Tanner (cct@faith.asd.sgi.com) Silicon Graphics - Advanced Graphics Division _____________________________________________________________ From guest Thu Aug 18 08:26:25 1994 From: "Shahram Shiri" Date: Thu, 18 Aug 1994 11:26:18 -0400 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: info-performer@sgi.sgi.com Subject: Perfly .ptu core dump on Onyx Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 To: info-performer@sgi.com Last week I tried the .ptu formats suggested by Chris Tanner. The program perfly coke.ptu works fine on an Indigo-2/Extreme. However, the same program on an Onyx crashes. Tracing the problem, core dump occurs inside the ilggettile.c, ilgettile() function. Apparently, it should return a pointer to the memory location where imagevision library saves the image. On Indigo2/Extreme, the pointer to the memory comes back fine, however, on Onyx, there is no return. I commented out the ilHwAccelerate() function in ilgettile.c and the problem still exists. Anyone have had similar problems? I tried the program on two different Onyxes and it failed on both systems. -Ron -- ------------------------------------------------------------------------- |Shahram Shiri/ ron@leaf.gsfc.nasa.gov Tel: 301-286-9985 | |NCCS/Scientific Visualization Studio Fax: 301-286-1634 | |NASA/Goddard Space Flight Center, Hughes/STX | |Greenbelt, Maryland 20771 | ------------------------------------------------------------------------- From guest Thu Aug 18 09:19:26 1994 Date: Thu, 18 Aug 94 18:19 MDT From: matomira@lig.di.epfl.ch (Fernando D. Mato Mira) To: info-performer@sgi.sgi.com Subject: modelers Reply-To: matomira@epfl.ch Hello, Can somebody tell me which are the pros and cons of Multigen with respect to DWB? I also need the corresponding email adresses to get pricing info and so on. Thanks Fernando D. Mato Mira Computer Graphics Lab Swiss Federal Institute of Technology (EPFL) Phone : +41 (21) 693 - 5248 CH-1015 Lausanne FAX : +41 (21) 693 - 5328 Switzerland E-mail : matomira@di.epfl.ch From guest Thu Aug 18 09:34:47 1994 From: "Darin C. Partridge" Date: Thu, 18 Aug 1994 10:34:32 -0600 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: info-performer@sgi.sgi.com Subject: Kinematics Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 I know I'm shouldn't be asking this non-performer question but you may be able to help me. I'm building a modeler and have implemented kinematics. The only problem is that my inverse kinematics is giving me problems. Does anyone know where I could go to get some source or something to give me a clue as to where I went wrong? Thanks for any info, Darin (darin@chaos.idec.sdl.usu.edu) From guest Thu Aug 18 09:54:07 1994 From: "Michael Jones" Date: Thu, 18 Aug 1994 09:53:54 -0700 In-Reply-To: matomira@lig.di.epfl.ch (Fernando D. Mato Mira) "modelers" (Aug 18, 6:19pm) References: X-Mailer: Z-Mail (3.2a.627 27jun94 MediaMail) To: matomira@epfl.ch, info-performer@sgi.sgi.com Subject: Re: modelers Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii On Aug 18, 6:19pm, Fernando D. Mato Mira wrote: > Subject: modelers : Can somebody tell me which are the pros and cons of :Multigen with respect to DWB? : I also need the corresponding email adresses to get :pricing info and so on. Engineers representing both companies are sure to post a response to this question, so I'll leave the answer to them. I did want to point out that a few other modeling tools also exist that you might want to check out if your needs are not as advanced: Medit, by Medit Productions 100346.1122@compuserve.com Seems very nice to me. It's simpler than the two that you mention, but might be sufficient for your needs. I3DM, by John Kichury I even simpler yet, but is the lowest cost of all in that it's included free in the Friends of Performer gifts distributed with IRIS Performer 1.2 in the "/usr/src/Performer/friends" directory. All four of these tools (and several others) work well with IRIS Performer. -- Be seeing you, Phone:415.390.1455 Fax:415.390.2658 M/S:8U-590 Michael T. Jones Silicon Graphics, Advanced Graphics Division mtj@sgi.com 2011 N. Shoreline Blvd., Mtn. View, CA 94039-7311 From guest Thu Aug 18 10:03:52 1994 From: tidrowd@cc.tacom.army.mil Date: Thu, 18 Aug 94 13:01:12 EDT Encoding: 822 Text To: info-performer@sgi.sgi.com Subject: Multi-Channel Option (MCO) problems This isn't exactly a Performer question, but I figure a fair number of Performer users also use MCO to get multiple displays. Anyway, here's the question: Does anyone know of a GL command, or a combination of GL commands, that will crash the graphics system on an Onyx (pipe 1 of a two pipe system) but not on a SkyWriter RealityEngine system (pipe 0 this time)??? We have a third-party application that is causing this behavior and have not had much success in solving the problem. Any help will be greatly appreciated. Don Tidrow Elect. Engineer US Army Tank-Automotive Command From guest Thu Aug 18 10:21:23 1994 From: "Michael Jones" Date: Thu, 18 Aug 1994 10:21:01 -0700 In-Reply-To: "Darin C. Partridge" "Kinematics" (Aug 18, 10:34am) References: <9408181034.ZM1955@paradox.idec.sdl.usu.edu> X-Mailer: Z-Mail (3.2a.627 27jun94 MediaMail) To: "Darin C. Partridge" , info-performer@sgi.sgi.com Subject: Re: Kinematics Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii On Aug 18, 10:34am, Darin C. Partridge wrote: > Subject: Kinematics :I know I'm shouldn't be asking this non-performer question but you may :be able to help me. Why not ask? I'm always curious to know what people are doing with IRIS Performer. How else can we know how to prioritize the many new features that we're planning. :I'm building a modeler and have implemented kinematics. The only :problem is that my inverse kinematics is giving me problems. Is this modeler something that will be made public? :Does anyone know where I could go to get some source or something :to give me a clue as to where I went wrong? Here are a few references that you might consult. You've probably already read these, but if you have missed any, you'll want to see how others approached the problem. As Isaac Newton said, "If I have seen further than others, it is because I have stood on the shoulders of giants." [If you went to the SIGGRAPH course #14 that several of us in the IRIS Performer group spoke at, you'll recognize the quote -- sorry for the repitition.] @InProceedings{kuder-1992-using, author = "Karen Kuder", title = "Using inverse kinematics to position articulated figures", pages = "121", journal = "Proceedings of the 1992 Western Computer Graphics Symposium" , year = "1992", month = "April", conference = "held in Banff, Alberta; 6-8 April 1992", keywords = "", annote = "", } @InProceedings{welman-1992-inverse, author = "Chris Welman and Tom Calvert", title = "Inverse kinematics through differential manipulation", pages = "123--128", journal = "Proceedings of the 1992 Western Computer Graphics Symposium" , year = "1992", month = "April", conference = "held in Banff, Alberta; 6-8 April 1992", keywords = "", annote = "", } @InProceedings{badler-1991-interactive, author = "Cary B. Phillips nad Norman I. Badler", title = "Interactive behaviors for bipedal articulated figures", pages = "359--362", journal = "Computer Graphics (SIGGRAPH '91 Proceedings)", volume = "25", number = "4", year = "1991", month = "July", editor = "Thomas W. Sederberg", conference = "held in Las Vegas, Nevada; 28 July - 2 August 1991", keywords = "interactive manipulation, inverse kinematics, articulated figures, balance, behavioral animation" , annote = "", } @InBook{girard-1991-constrained, author = "Michael Girard", title = "Constrained optimization of articulated animal movement in computer animation" , pages = "209--232", booktitle = "Making them move: mechanics, control, and animation of articulated figures" , year = "1991", editor = "Norman I. Badler and Brian A. Barsky and David Zeltzer", publisher = "Morgan Kaufmann", keywords = "animation, forward kinematics, inverse kinematics, dynamics" , annote = "", } @Article{monheit-1991-kinematic, author = "Gary Monheit and Norman I. Badler", title = "A kinematic model of the human spine and torso", pages = "29--38", journal = "IEEE Computer Graphics and Applications", volume = "11", number = "2", month = "March", year = "1991", keywords = "animation, inverse kinematics, motion constraints", annote = "By modeling a flexible torso and spine, they produce motion of human bodies more realistic. " , } @InProceedings{rosebush-1991-kinematics, author = "Judson Rosebush", title = "Kinematics for animators and graphicists", pages = "1--16", booktitle = "COMPUGRAPHICS '91", volume = "II", year = "1991", conference = "held in Sesimbra, Portugal; 16-20 September 1991", keywords = "animation", annote = "", } @InProceedings{welman-1991-kinematic, author = "Chris Welman", title = "Kinematic constraints for articulated figure manipulation", pages = "44--49", journal = "Proceedings of the 1991 Western Computer Graphics Symposium" , year = "1991", month = "April", conference = "held in Vernon, B.C.; 8-10 April 1991", keywords = "forward kinematics, inverse kinematics", annote = "", } @Article{boulic-1990-global, author = "Ronan Boulic and Nadia Magnenat Thalmann and Daniel Thalmann" , title = "A global human walking model with real-time kinematic personification" , journal = "The Visual Computer", pages = "344--358", volume = "6", number = "6", month = "December", year = "1990", keywords = "animation, biomechanics, gait, inverse kinematics", annote = "", } @InProceedings{boulic-1990-human, author = "Ronan Boulic and Nadia Magnenat-Thalmann and Daniel Thalmann" , title = "Human free-walking model for a real-time interactive design of gaits" , pages = "61--79", booktitle = "Computer Animation '90 (Second workshop on Computer Animation)" , year = "1990", month = "April", editor = "N. Magnenat-Thalmann and D. Thalmann", publisher = "Springer-Verlag", conference = "held in Geneva, Switzerland; 25-27 April 1990", keywords = "animation, inverse kinematics, biomechanics, personification, behaviour, prediction, correction" , annote = "", } @InProceedings{bruderlin-1989-goal-directed, author = "Armin Bruderlin and Thomas W. Calvert", title = "Goal-Directed, Dynamic Animation of Human Walking", pages = "233-242", journal = "Computer Graphics (SIGGRAPH '89 Proceedings)", volume = "23", number = "3", year = "1989", month = "July", editor = "Jeffrey Lane", conference = "held in Boston, Massachusetts; 31 July -- 4 August 1989", keywords = "animation, goal-directed animation, human figure animation, dynamics, kinematics, inverse kinematics" , annote = "", } @InProceedings{chadwick-1989-layered, author = "John E. Chadwick and David R. Haumann and Richard E. Parent" , title = "Layered Construction for Deformable Animated Characters", pages = "243-252", journal = "Computer Graphics (SIGGRAPH '89 Proceedings)", volume = "23", number = "3", year = "1989", month = "July", editor = "Jeffrey Lane", conference = "held in Boston, Massachusetts; 31 July -- 4 August 1989", keywords = "free form deformations, robotic manipulators, kinematics, dynamics, character animation" , annote = "", } @InProceedings{calvert-1988-the, author = "Tom Calvert", title = "The challenge of human figure animation", pages = "203--210", journal = "Proceedings of Graphics Interface '88", year = "1988", month = "June", conference = "held in Edmonton, Alberta; 6-10 June 1988", keywords = "human movement, kinematics, dynamics, expert systems, movement representation, modelling" , annote = "", } @InProceedings{badler-1986-multi-dimensional, author = "Norman I. Badler and Kamran H. Manoochehri and David Baraff" , title = "Multi-dimensional input techniques and articulated figure positioning by multiple constraints" , pages = "151--169", booktitle = "Proceedings of 1986 Workshop on Interactive 3D Graphics", year = "1986", month = "October", editor = "Frank Crow and Stephen M. Pizer", conference = "held in Chapel Hill, North Carolina; 23-24 October 1986", keywords = "six degree input device, kinematics", annote = "", } @InProceedings{elmaraghy-1986-kinematic, author = "Hoda A. ElMaraghy", title = "Kinematic and geometric modelling and animation of robots", pages = "15--19", journal = "Proceedings of Graphics Interface '86", year = "1986", month = "May", editor = "M. Green", conference = "held in Vancouver, B.C.; 26-30 May 1986", keywords = "robot kinematics, graphical simulation and animation, robot off-line programming" , annote = "", } @Article{medland-1986-the, author = "A. J. Medland", title = "The alignment of associated space in an automated {CAD}-kinematics procedure" , @Article{medland-1986-the, author = "A. J. Medland", title = "The alignment of associated space in an automated {CAD}-kinematics procedure" , pages = "239--244", journal = "Computers and Graphics", volume = "10", number = "3", year = "1986", keywords = "", annote = "", } @InProceedings{leu-1984-computer, author = "M. C. Leu and R. Mahajan", title = "Computer Graphic Simulation of Robot Kinematics and Dynamics" , booktitle = "Robots 8 Conference Proceeding (2 vold)", pages = "4/80--101", volume = "1", year = "1984", organization = "Robotics Int. Sme", conference = "held in Detroit, MI; 4--7 June 1984", keywords = "I34 simulation, I3m robot kinematics, I3m robot dynamics", } @Article{leu-1984-simulation, author = "M. C. Leu and R. Mahajan", title = "Simulation of Robot Kinematics Using Interactive Computer Graphics" , year = "1984", month = "January", journal = "COED (USA)", volume = "4", number = "1", pages = "11-14", keywords = "I3m robot kinematics and I34 simulation and I32 interactive systems" , } @Article{adams-1981-kinematics, author = "J. A. Adams", title = "Kinematics with Computer Graphics", year = "1981", journal = "Comput. and Educ.", volume = "5", pages = "245-264", keywords = "I3m kinematics", } @InProceedings{parke-1978-interactive, author = "F. I. Parke and M. Friedell", title = "Interactive simulation of biomechanical systems: the kinematics and stress of the human knee" , booktitle = "ACM 78 Proc. of the Annual Conf.", pages = "759--764", year = "1978", keywords = "biomechanical system, interactive simulation", } -- Be seeing you, Phone:415.390.1455 Fax:415.390.2658 M/S:8U-590 Michael T. Jones Silicon Graphics, Advanced Graphics Division mtj@sgi.com 2011 N. Shoreline Blvd., Mtn. View, CA 94039-7311 From guest Thu Aug 18 10:33:54 1994 From: "Michael Jones" Date: Thu, 18 Aug 1994 10:33:41 -0700 In-Reply-To: tidrowd@cc.tacom.army.mil "Multi-Channel Option (MCO) problems" (Aug 18, 1:01pm) References: <9407187772.AA777240072@tacom-ccmail4.local> X-Mailer: Z-Mail (3.2a.627 27jun94 MediaMail) To: tidrowd@cc.tacom.army.mil, info-performer@sgi.sgi.com Subject: Re: Multi-Channel Option (MCO) problems Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii On Aug 18, 1:01pm, tidrowd@cc.tacom.army.mil wrote: > Subject: Multi-Channel Option (MCO) problems : : This isn't exactly a Performer question, but I figure a fair number of : Performer users also use MCO to get multiple displays. Anyway, here's : the question: Does anyone know of a GL command, or a combination of GL : commands, that will crash the graphics system on an Onyx (pipe 1 of a : two pipe system) but not on a SkyWriter RealityEngine system (pipe 0 : this time)??? We have a third-party application that is causing this : behavior and have not had much success in solving the problem. Any : help will be greatly appreciated. Call the Hot Line. Have the last few lines of your "/usr/adm/SYSLOG" file handy, know the exact installed hardware and software versions, and if at all possible, be prepared to ftp the core dump file. This is my only advice without more information. -- Be seeing you, Phone:415.390.1455 Fax:415.390.2658 M/S:8U-590 Michael T. Jones Silicon Graphics, Advanced Graphics Division mtj@sgi.com 2011 N. Shoreline Blvd., Mtn. View, CA 94039-7311 From guest Thu Aug 18 12:17:58 1994 From: davec@cory.coryphaeus.com (David Cooper) Subject: Re: modelers To: info-performer@sgi.sgi.com Date: Thu, 18 Aug 1994 10:53:32 -0800 (PDT) In-Reply-To: from "Fernando D. Mato Mira" at Aug 18, 94 06:19:00 pm X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 656 > > > Hello, > > Can somebody tell me which are the pros and cons of > Multigen with respect to DWB? > I also need the corresponding email adresses to get > pricing info and so on. > To avoid turning the mailing list into a slagging match between DWB and Multigen this has been answered direct to the originator. If anyone out there has any questions regarding DWB or any coryphaeus product email me at davec@coryphaeus.com for any information -- Dave Cooper Coryphaeus Software, Inc. davec@coryphaeus.com 985 University Ave., Suite 31, Los Gatos, CA 95030 Tel: 408-395-4537 Fax: 408-395-6351 From guest Thu Aug 18 12:34:04 1994 (5.65c/IDA-1.4.4 for ); Thu, 18 Aug 1994 15:33:54 -0400 Date: Thu, 18 Aug 1994 15:33:54 -0400 From: John C Vanderburgh To: info-performer@sgi.sgi.com Subject: dwb flames files... Hey all... this isn't exactly a Performer problem but... I got the "flames" files from Dave C at Coryphaeus (that were discussed in previous postings) and am having problems... I receive my mail on a Sun system, so I extracted the uuencoded files via the "mailtool" program. I shipped everything (via ftp) to my SGI box and tried uncompressing it (after uudecoding it) but got the message "uncompress: flames.tar.Z: corrupt input." So, I uncompressed it on the Sun box, and then tried to un-tar it, but got a "checksum error". Has anyone else had problems like these??? (I had Dave send me the files again and I got the same result, so I'm assuming the problem is on my end.) Thanks... John Vanderburgh Air Force Inst of Technology WPAFB, OH From guest Thu Aug 18 19:37:31 1994 From: "Chris Tanner" Date: Thu, 18 Aug 1994 19:37:27 -0700 In-Reply-To: matomira@lig.di.epfl.ch (Fernando D. Mato Mira) "texgen" (Aug 18, 2:42pm) References: X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: matomira@epfl.ch Subject: Re: texgen Cc: info-performer@faith Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 On Aug 18, 2:42pm, Fernando D. Mato Mira wrote: > Subject: texgen > > Hello, > > in info-performer you wrote: > > >>> > long pfuPreDrawLinearMap(pfTraverser *trav, void *data) > { > pfDataPool *dp; > float *paramsX, *paramsY; > > dp = pfAttachDPool(CallBacks); > if (dp == NULL) > pfuTexgenParams(NULL, NULL); > paramsX = (float *)pfDPoolFind(dp, 1); > paramsY = (float *)pfDPoolFind(dp, 2); > (trav, trav); > (data, data); > > texgen(TX_S, TG_SPHEREMAP, paramsX); > texgen(TX_T, TG_SPHEREMAP, paramsY); > texgen(TX_S, TG_ON, NULL); > texgen(TX_T, TG_ON, NULL); > > return NULL; > } > >>> > > Shouldn't this be TG_LINEAR? Cut&paste syndrome, probably. > > Thanks, > > Fernando D. Mato Mira > Computer Graphics Lab > Swiss Federal Institute of Technology (EPFL) Phone : +41 (21) 693 - 5248 > CH-1015 Lausanne FAX : +41 (21) 693 - 5328 > Switzerland E-mail : matomira@di.epfl.ch >-- End of excerpt from Fernando D. Mato Mira Exactly... Sorry 'bout that... Later, Chris -- _____________________________________________________________ Chris Tanner (cct@faith.asd.sgi.com) Silicon Graphics - Advanced Graphics Division _____________________________________________________________ From guest Fri Aug 19 07:25:36 1994 Date: Fri, 19 Aug 94 16:25 MDT From: matomira@lig.di.epfl.ch (Fernando D. Mato Mira) To: info-performer@sgi.sgi.com Subject: state stack Reply-To: matomira@epfl.ch Hello, Is there a way of altering the current material without doing pfApplyMtl, but by calling lmdef(DEFMATERIAL ... If so, a) How do I get the current material index (to use in the lmdef) ? b) Will this get restored after subsequent pfPushState/ pfPopState (I doubt it). Otherwise, if do: pfPushState(); mat = pfGetCurMtl(side); modify mat pfApplyMtl(mat); Will this alter the previous entry in the stack or will it really create a new material context? The reason why is ask this is because I want to avoid allocating/copying /modifying/ref checking/returning to material pool materials to be used in draw callbacks (because of pipe latency, if the material associated with a node (interior node, I use callbacks) gets modified in the app, I can't just modify the structure). I have closures (lisp), so it's easy for me to pass a callback that just looks into some variables (instead of having to pass a `data' block). Even if I passed some data block, avoiding using a pfMaterial means I do not have to check in the post callback if the material is different, checking ref (several callbacks could use the same pfMaterial) and freeing, but just letting the lisp garbage collector take care of freeing the closure (OK, I still have to keep a refcount on closures and check on the post, because it could happen that the callback for a node is changed, this was the only one to have it, and there's still a frame down the pipe (in the C side) that needs it). Also, like this a node could also alter one property (let's say, front material specular), although if I have to do pfApplyMtl(mat); I do not gain anything there, but if the attribute is highly animated, I do not need to go copying full materials every frame, at least. P.S. : If anybody sent me some info on Modelgen or DWB, or wrote it to this list, I didn't get anything, I got some Mail problem today (/usr/tmp full, Mail ate my mailbox, no mailer-daemon warnings to the world). Thanks, Fernando D. Mato Mira Computer Graphics Lab Swiss Federal Institute of Technology (EPFL) Phone : +41 (21) 693 - 5248 CH-1015 Lausanne FAX : +41 (21) 693 - 5328 Switzerland E-mail : matomira@di.epfl.ch From guest Fri Aug 19 11:16:42 1994 From: bohrer@netcom.com (Jan) Subject: Changing colors of groups of polygons To: info-performer@sgi.sgi.com Date: Fri, 19 Aug 1994 11:16:58 -0700 (PDT) X-Mailer: ELM [version 2.4 PL23] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 2244 Hi - I'm working on dynamically changing colors of groups of polygons on a real time model and having some problems. I wrote some test code that sets several different polygon groups to different colors while not changing the rest of the model. When I execute the program, the polygons are basically all the same color scheme and and some of the colors change intermitantly (they flash to white and then back to whatever color was there before). Each polygon group is supposed to be a different color but they all seem to come back with the same color scheme. Each polygon group is identified with a Geode at load time through a callback function. I have tried the pfGSetAttr with different attribute bindings (PFGS_OVERALL, PFGS_PER_PRIM, and PFGS_PER_VERTEX). With the OVERALL flag set the entire model take on one color and then just flashes back and forth. By PER_PRIM it appears as if individual polygons within the Geode take on different colors, and with PER_VERTEX, I get a psychodelic model that also flashes. I am including a fragment of code where I am actually attempting to change the color. I would appreciate any suggestions for help with this problem. TIA, Jan please note that facet is a pointer to a structure which contains the facet (polygon group) id number and the GeoSet pointer. This facet number is used to feed color change data to the polygons in real-time from another process. Also, the color setting occurs in the postframe processing at approximately 20 times per second. **************************** Code Fragment Included ************************* if (facet->facet_id == 9) { red = 1.0; green = 0.0; blue = 0.0; pfSetVec4(color, red, green, blue, alpha); pfGSetAttr(facet->facet_geoset, PFGS_COLOR4, PFGS_PER_PRIM, color, NULL); } pfDCSRot(model, h, p, r); -- ___ ___....-----'---`-----....___ ========================================= Jan Bohrer ___`---..._______...---'___ Klutz Extraordinaire (___) _|_|_|_ (___) Can We Go Skiing Now? \\____.-'_.---._`-.____// ~~~~`.__`---'__.'~~~~ `~~~' He's Dead Jim... From guest Fri Aug 19 12:07:28 1994 Date: Fri, 19 Aug 94 15:05:07 -0700 From: laser@ice.bailey.com (Mark A Lasersohn) Apparently-To: info-performer@sgi.sgi.com Content-Length: 505 Hello, I am new to Performer - so I have a pretty simple question. I have been trying to get Inventor .iv files to load into perfly. I've tried both 1.0 and 2.0 iv files and usually - though not always, perfly tells me it can't load them (after finding and loading the textures declared therein). Can anyone tell me how it is discriminating? Mark Lasersohn Bailey Controls Company laser@bailey.com From guest Fri Aug 19 16:51:22 1994 ; Fri, 19 Aug 1994 19:51:03 -0400 id AA01809; Fri, 19 Aug 1994 16:55:14 PST Organization: MultiGen Inc X-Charset: MACINTOSH X-Umcp-To: INFO PERFORMER From: Marcus To: INFO PERFORMER Date: Fri, 19 Aug 1994 16:37:48 PST Subject: Re: 2-color directional ligh Reply to: RE>2-color directional lightpo I've learned (thanks Paul Levy) that unidirectional pfLightPoints, when placed back to back, cannot be made to form a single seamless (2 color) light. After some investigation, I've found that Performer clamps the vertical and horizontal envelopes to 179.0 degrees maximum. This leaves a 2 degree gap between each light when viewed from the side, wherein both lights are invisable for a brief moment. Shouldn't the envelopes be unrestricted, at least within 180.0 degrees if not 360.0? Regards, Marcus Barnes, Member Technical Staff MultiGen Inc., 1884 The Alameda, San Jose CA, 95126 PH: (408) 261 4110 FX: (408) 247 4329 EMAIL: multigen!marcus@uunet.UU.NET From guest Sun Aug 21 22:13:35 1994 From: "David Pratt" Date: Sun, 21 Aug 1994 22:15:30 -0700 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: info-performer@sgi.sgi.com Subject: Problems with 1.2 on Indigo2 Extreme/XL Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Good morning, I hope you had a better weekend then I did. When I run perfly on my Indigo2 Extreme/XL (Two pipeline Indigo2, see the hinv/uname below) on the Extreme pipline it opens up two full screen windows. The "Welcome to Performer" screen appears, flashes black, then the "Welcome to Performer" screen comes back up. The first one does not have the gui widgets, the second one is ok. In order to get to the good one, I have to push (ALT-F3) the other one to the back. Suggestions? I can live with that, but the real problem is on the local code. On all the machines, but this one, it comes up ok. On mine, it does not display the lit terrain. I know the terrain is there, when I turn on texturing I can see it. I would feel better if the two problems were the same. I have recompiled the code to make sure it was not that. Any Ideas? Dave Here are the specifics of my system: medusa:/usr/work/pratt % hinv 1 150 MHZ IP22 Processor FPU: MIPS R4010 Floating Point Chip Revision: 0.0 CPU: MIPS R4400 Processor Chip Revision: 5.0 On-board serial ports: 2 On-board bi-directional parallel port Data cache size: 16 Kbytes Instruction cache size: 16 Kbytes Secondary unified instruction/data cache size: 1 Mbyte Main memory size: 96 Mbytes EISA bus: adapter 0 Iris Audio Processor: version A2 revision 1.1.0 Integral Ethernet: ec0, version 1 Disk drive: unit 7 on SCSI controller 1 Disk drive: unit 6 on SCSI controller 1 CDROM: unit 4 on SCSI controller 1 Tape drive: unit 3 on SCSI controller 1: DAT Integral SCSI controller 1: Version WD33C93B, revision D Disk drive / removable media: unit 2 on SCSI controller 0 Disk drive: unit 1 on SCSI controller 0 Integral SCSI controller 0: Version WD33C93B, revision D Graphics board: XL Graphics board: GU1-Extreme medusa:/usr/work/pratt % uname -a IRIX medusa 5.2 02282016 IP22 mips medusa:/usr/work/pratt % versions performer_dev I = Installed, R = Removed Name Date Description I performer_dev 08/08/94 IRIS Performer1.2/Irix5 Development Software I performer_dev.man 08/08/94 IRIS Performer1.2/Irix5 Documentation I performer_dev.man.relnotes 08/08/94 IRIS Performer1.2/Irix5 Release Notes I performer_dev.src 08/08/94 IRIS Performer1.2/Irix5 Sample Code I performer_dev.src.pguide 08/08/94 IRIS Performer1.2/Irix5 Programmers Guide Examples I performer_dev.src.sample 08/08/94 IRIS Performer1.2/Irix5 Sample Application I performer_dev.sw 08/08/94 IRIS Performer1.2/Irix5 Development Environment I performer_dev.sw.debug 08/08/94 IRIS Performer1.2/Irix5 Debug Libraries I performer_dev.sw.performer 08/08/94 IRIS Performer1.2/Irix5 Headers I performer_dev.sw.pfdwb 08/08/94 IRIS Performer1.2/Irix5 Sample DWB File Conversion Library I performer_dev.sw.pfflt 08/08/94 IRIS Performer1.2/Irix5 Sample Flight File Conversion Library I performer_dev.sw.pfsgi 08/08/94 IRIS Performer1.2/Irix5 Sample SGI File Conversion Library I performer_dev.sw.pfutil 08/08/94 IRIS Performer1.2/Irix5 Utility Library I performer_dev.sw.static 08/08/94 IRIS Performer1.2/Irix5 Non-Shared Libraries medusa:/usr/work/pratt % From guest Mon Aug 22 10:22:53 1994 Date: Mon, 22 Aug 1994 12:19:01 -0500 (CDT) From: Olukayode B Olofinboba Subject: John Burwell or IRIS performer SGI contact To: info-performer@sgi.sgi.com Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Does anyone know the correct email address for John Burwell? He's mentioned in the Performer FAQ as the contact for product info about performer or SGI visual sim. issues. I tried johnnyb@asd.sgi.com and the mail was returned. Thanks, Olu Olofinboba olofi001@maroon.tc.umn.edu From guest Mon Aug 22 11:07:23 1994 Date: Mon, 22 Aug 1994 18:07:48 GMT From: Kevin Russo To: info-performer@sgi.sgi.com Subject: pfuFindHash broken Fellow Performers, I think there's a bug in pfuFindHash() (Perf 1.2). It should set the key before trying to find anything. elt->key = getKey(ht, elt); After I made this fix, it worked, but then my app ran like a DOG. Looking closer, it occurred to me that pfuFindHash is doing a linear probe on a hash table!!? I got vastly improved performance by taking the pfuRemoveHash algorithm and stripping out the "remove" part. Does this sound right? Kevin -- Kevin Russo russo@ait.nrl.navy.mil US Naval Research Lab Washington, DC From guest Mon Aug 22 17:18:39 1994 (5.67b8/IDA-1.5 for info-performer@sgi.com); Tue, 23 Aug 1994 10:17:51 +1000 (5.65c/IDA-1.5); Tue, 23 Aug 1994 10:12:13 +1000 (5.65c/IDA-1.4.4); Tue, 23 Aug 1994 10:06:57 +1000 Date: Tue, 23 Aug 1994 10:06:04 +1000 (EST) From: Simon Bennett Subject: Re: dwb flames files... To: John C Vanderburgh Cc: info-performer@sgi.sgi.com In-Reply-To: <199408181933.AA26293@stealth.afit.af.mil> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Thu, 18 Aug 1994, John C Vanderburgh wrote: > Hey all... this isn't exactly a Performer problem but... > I got the "flames" files from Dave C at Coryphaeus (that were discussed in previous postings) and am having problems... > I receive my mail on a Sun system, so I extracted the uuencoded files via the "mailtool" program. I shipped everything (via ftp) to my SGI box and tried uncompressing it (after uudecoding it) but got the message "uncompress: flames.tar.Z: corrupt input." So, I uncompressed it on the Sun box, and then tried to un-tar it, but got a "checksum error". Has anyone else had problems like these??? (I had Dave send me the files again and I got the same result, so I'm assuming the problem is on my end.) I've also got Dave's flame effect and had no problems. Built fine and looks great. I'd say the problem is definitely at your end... Trying skipping the "mailtool" and extracting the parts by hand... +--------------------------------------------------------------------------------+ Simon Bennett simonb@wormald.com.au Wormald Technology Advanced Systems Engineering Ph: +61 2 981 0611 (x512) "Business meetings are really very useful... For one thing they clearly demonstrate just how many people the company can do *without*" From guest Mon Aug 22 18:07:11 1994 (5.65c/IDA-1.4.4 for ); Mon, 22 Aug 1994 21:07:04 -0400 Date: Mon, 22 Aug 1994 21:07:04 -0400 From: John C Vanderburgh To: info-performer@sgi.sgi.com Subject: dwb flames... Hey all, I previously posted a blurb about some problems I was having extracting some "dwb" files received from Dave at Coryphaeus. Thanks to some replies, I discovered that the problem was some blank lines in the resultant "cat" file from the four e-mailed parts. I took the blank lines out and everything worked like clockwork... By the way, great demo!! Thanks for all the replies, John Vanderburgh Air Force Inst of Technology WPAFB, OH From guest Tue Aug 23 00:40:59 1994 From: norbert@shark.paris.sgi.com (Norbert BIANCHIN) Date: Tue, 23 Aug 1994 08:41:14 -0600 In-Reply-To: Olukayode B Olofinboba "John Burwell or IRIS performer SGI contact" (Aug 22, 12:19pm) References: X-Face: 1mi: {7!PS}0`(6{{9lNvh:.=$m&.hb5&kQR'-=yEP>d+VL_oaf)sTzR]w9y2]=g6lN*<*h2K'e\+f;,wPeU_V!,8XYT*WHzZT,.t2^ld]QQ^DA'{#m}0j$o1$)X2HI^FAi7b)]X9Io'==q?!i>y. ?x&kp:Vp;|=CO:{Ir(zP\O[Iid+~$!)8RU>txetK)FHi`9`f:$xCTElo~#\Wtmv Subject: John Burwell or IRIS performer SGI contact [ Does anyone know the correct email address for John Burwell? [ He's mentioned in the Performer FAQ as the contact for product [ info about performer or SGI visual sim. issues. I tried [ johnnyb@asd.sgi.com and the mail was returned. [ [ Thanks, [ [ Olu Olofinboba [ olofi001@maroon.tc.umn.edu [ [ [ [ >-- End of excerpt from Olukayode B Olofinboba Try : jburwell@asd.sgi.com JOHN BURWELL Good luck Norbert -- +----------------------------------------------------------------+ | NN NN BBBBBB Country ID : IFR-340 | | NNN NN BB BB SGI Office Location : PARIS | | NNNN NN BBBBBB Name : Norbert BIANCHIN | | NN NNNN BB BB | | NN NNN BB BB E-mail Address : norbert@paris.sgi.com | | NN NN BBBBBB Fax Number : 33 1 34 88 80 76 | +----------------------------------------------------------------+ From guest Tue Aug 23 02:22:11 1994 Resent-Message-Id: <199408230922.CAA22501@sgi.sgi.com> Via: uk.ac.edinburgh.festival; Tue, 23 Aug 1994 10:20:16 +0100 via SMTP (940627.SGI.8.6.9/910110.SGI) id AAA18463; Tue, 23 Aug 1994 00:58:52 -0700 for @sgi.sgi.com:performers@vr1.engin.umich.edu id AA00887; Tue, 23 Aug 94 00:41:00 -0700 via SMTP (931110.SGI/911001.SGI) for /usr/lib/sendmail info-performer-dist id AA00884; Tue, 23 Aug 94 00:40:59 -0700 via SMTP (920330.SGI/920502.SGI) for info-performer@holodeck.asd.sgi.com id AA18571; Mon, 22 Aug 94 23:42:00 -0700 via SMTP (940627.SGI.8.6.9/910110.SGI) for id XAA13057; Mon, 22 Aug 1994 23:41:55 -0700 via SMTP (920330.SGI/920502.SGI) for info-performer@sgi.sgi.com id AA22337; Mon, 22 Aug 94 23:41:44 -0700 for info-performer@sgi.sgi.com id AA26758; Tue, 23 Aug 94 08:41:14 +0200 From: Norbert BIANCHIN Date: Tue, 23 Aug 1994 08:41:14 -0600 In-Reply-To: Olukayode B Olofinboba "John Burwell or IRIS performer SGI contact" (Aug 22, 12:19pm) References: X-Face: 1mi: {7!PS}0`(6{{9lNvh:.=$m&.hb5&kQR'-=yEP>d+VL_oaf)sTzR]w9y2]=g6lN*<*h2K'e\+f;,wPeU_V!,8XYT*WHzZT,.t2^ld]QQ^DA'{#m}0j$o1$)X2HI^FAi7b)]X9Io} X-Mailer: Z-Mail-SGI (3.0S.1023 23oct93 MediaMail) To: olofi001@maroon.tc.umn.edu, info-performer@sgi.sgi.com Subject: Re: John Burwell or IRIS performer SGI contact Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Resent-Date: Tue, 23 Aug 94 10:19:57 BST Resent-From: mmdf@festival.edinburgh.ac.uk Resent-To: olofi001@maroon.tc.umn.edu Resent-To: info-performer@sgi.sgi.com Sender: mmdf@festival.edinburgh.ac.uk On Aug 22, 12:19pm, Olukayode B Olofinboba wrote: > Subject: John Burwell or IRIS performer SGI contact [ Does anyone know the correct email address for John Burwell? [ He's mentioned in the Performer FAQ as the contact for product [ info about performer or SGI visual sim. issues. I tried [ johnnyb@asd.sgi.com and the mail was returned. [ [ Thanks, [ [ Olu Olofinboba [ olofi001@maroon.tc.umn.edu [ [ [ [ >-- End of excerpt from Olukayode B Olofinboba Try : jburwell@asd.sgi.com JOHN BURWELL Good luck Norbert -- +----------------------------------------------------------------+ | NN NN BBBBBB Country ID : IFR-340 | | NNN NN BB BB SGI Office Location : PARIS | | NNNN NN BBBBBB Name : Norbert BIANCHIN | | NN NNNN BB BB | | NN NNN BB BB E-mail Address : norbert@paris.sgi.com | | NN NN BBBBBB Fax Number : 33 1 34 88 80 76 | +----------------------------------------------------------------+ From guest Tue Aug 23 02:41:50 1994 Date: Tue, 23 Aug 94 11:39:22 +0200 From: gce@scl.syseca.fr (Cedric Gautier ) To: info-performer@sgi.sgi.com Subject: Performer with ImageVision, Movie and Video libraries ... Steve Hodecker LORAL Federal Systems Flight Systems Engineering hodeckes@VNET.IBM.COM wrotes .... First, has anyone used the ImageVision Library with Performer. Please comment on any issues that must be addressed to do this. .... Answer is: Bonjour ... Hello ... I'm using ImageVision, MovieLibrary and VideoLibrary mixed with Performer1.2. IL revision is 2.3, MV and VL revision are 2.0 running on IRIX 5.2 on Indy. All of this is used to mix live video and/or movie to performer animation taking care of camera lens deformation (IL is used to warp back the entry video or movie to non deformed images according to the camera lens, focal and zoom ...) The only problem I had using all of this staff was to be forced to handle the background image display through GL and not through ILdisplay because of one bug which cannot allows you to display background and foreground correctly ... (ILdisplay was perfectly running under IL 2.0 and IRIX 4.05 and Performer 1.2 for irix 4.05 but Movie and Video libraries were not available under IRIX 4.05) ... The link line is quite long ... with: pfsgi pfutil pf pr cil ilSGI il image movie dmedia cl awareaudio vl mpc gl audio X11 fm m fpe C but you may not need the DigitalMedia part of it ... I'll be glad to exchange information about this massive libraries mix for futur problems may be ! ... A bientot ... Cedric Cedric from Dahouet harbor on Armor coast in France - email: gce@syseca.fr ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ From guest Tue Aug 23 03:01:10 1994 (5.67b8/IDA-1.5 for info-performer@sgi.com); Tue, 23 Aug 1994 19:16:52 +1000 (5.65c/IDA-1.5 for ); Tue, 23 Aug 1994 18:07:43 +1000 (5.65c/IDA-1.4.4 for ); Tue, 23 Aug 1994 18:04:57 +1000 Date: Tue, 23 Aug 1994 18:04:36 +1000 (EST) From: Kathryn Loynes Sender: Kathryn Loynes Reply-To: Kathryn Loynes Subject: Associating node callbacks with pfClone To: Performer Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII Hi, I was wondering if it is possible to set up a node callback function which would be invoked during a pfClone traversal. I am cloning subsections of the scene graph within my application which represent dynamic models. Some of these models have articulated parts within them (represented by pfDCSs). After cloning I need to identify which new nodes represent the cloned pfDCSs. At present, after calling pfClone I am using pfuTraverse (with callback) to search through the new scene graph to identify the new pfDCSs. This obviously means two traversals. This could be reduced to one traversal if it was possible to set up node callback functions in the same way as pfNodeTravFuncs can be used to set up node callbacks for the major intersection, cull & draw traversals. I can't see any obvious way of doing this, if anyone has any ideas I'd be interested to hear them. Thanks in advance. ---------------------------------------------------------------------- Kathy Loynes | Wormald Technology kathyl@wormald.com.au | Advanced Systems Engineering | Ph: +61 2 981 0611 From guest Tue Aug 23 03:05:49 1994 Date: Tue, 23 Aug 94 12:01:35 +0200 From: gce@scl.syseca.fr (Cedric Gautier ) To: info-performer@sgi.sgi.com Subject: Inventor loader ... Hello, I am new to Performer - so I have a pretty simple question. I have been trying to get Inventor .iv files to load into perfly. I've tried both 1.0 and 2.0 iv files and usually - though not always, perfly tells me it can't load them (after finding and loading the textures declared therein). Can anyone tell me how it is discriminating? Mark Lasersohn Bailey Controls Company laser@bailey.com Answer is: try this version of the pfiv.c loader ... #include #include #include #include #include #include "pf.h" #include "pfsgi.h" #include "pfutil.h" /* case insensitive string equality test */ #define SAME(_a, _b) (strcasecmp(_a,_b) == 0) /* program capacity constants */ #define MAX_MATERIALS 512 #define MAX_TEXTURES 20000 #define MAX_STRING 1024 #define MAX_VERTICES 20000 #define MAX_NORMALS 20000 #define MAX_INDICES 20000 /* token types returned from parser */ #define T_END 0 #define T_KEYWORD 1 #define T_NUMBER 2 #define T_COMPLETE 3 /* attribute and mode definition */ typedef struct Style { pfGeoState *geostate; pfMaterial *material; pfTexture *texture; pfTexEnv *environment; long culling; } Style; /* function type and argument declarations */ static FILE *openFile(const char *fileName); static pfGeoState* defaultGeoState(void); static void processPolygons(pfGeode *geode, pfuBuilder *builder, Style *style); static int getToken (FILE *fp, char *token); static float ar = 0.5, ag = 0.5, ab = 0.5; static float dr = 0.5, dg = 0.5, db = 0.5; static float sr = 0.5, sg = 0.5, sb = 0.5; static float er = 0.5, eg = 0.5, eb = 0.5; static float shin= 0.0; extern pfNode * LoadIv (char *fileName, pfGeoState *geostate) { FILE *ivFile; pfuPoly polygon; pfuBuilder *builder = NULL; int tokenType = NULL; char token[256]; char verToken[256]; char texPath[256]; pfGeode *geode = NULL; pfNode *node = NULL; pfNode *hierarchy = NULL; int width = 0; int i; int j; long pending = 0; long numV = 0; long numT = 0; long numVI = 0; long numTI = 0; long numN = 0; long numNI = 0; long numTris = 0; long numSkip = 0; long numOther = 0; float alpha = 1.0f; float sx = 1.0f; float sy = 1.0f; float sz = 1.0f; double startTime = pfGetTime(); double elapsedTime = 0.0; static void *sharedArena = NULL; pfVec3 v[MAX_VERTICES]; pfVec3 n[MAX_NORMALS]; pfVec2 t[MAX_TEXTURES]; int vi[MAX_INDICES]; int ni[MAX_INDICES]; int f[MAX_INDICES]; int vCount = 0; int tCount = 0; int viCount = 0; int nCount = 0; int niCount = 0; int fCount = 0; Style style; strcpy(verToken, "point"); /* get pointer to IRIS Performer shared memory arena */ if (sharedArena == NULL) sharedArena = pfGetSharedArena(); /* open ".iv" file */ if ((ivFile = openFile(fileName)) == NULL) return NULL; /* give object a default geostate if nothing is passed in */ if (geostate == NULL) geostate = defaultGeoState(); /* the style structure contains the current modes and attributes */ style.geostate = geostate; style.material = NULL; style.culling = PFCF_BACK; /* give object a default texture environment */ style.texture = pfNewTex(sharedArena); style.environment = pfNewTEnv(sharedArena); pfTEnvMode(style.environment, PFTE_MODULATE); /* initialize utility library triangle/geoset builder */ builder = pfuNewBuilder(); /* read ".iv" file */ tokenType = getToken(ivFile, token); while (tokenType != T_END) { if (SAME(token, "ambientColor")) { getToken(ivFile, token); sscanf(token, "%f", &ar); getToken(ivFile, token); sscanf(token, "%f", &ag); getToken(ivFile, token); sscanf(token, "%f", &ab); tokenType = getToken(ivFile, token); } else if (SAME(token, "diffuseColor")) { getToken(ivFile, token); sscanf(token, "%f", &dr); getToken(ivFile, token); sscanf(token, "%f", &dg); getToken(ivFile, token); sscanf(token, "%f", &db); tokenType = getToken(ivFile, token); pfSetVec4(polygon.colors[0], dr, dg, db, alpha); } else if (SAME(token, "specularColor")) { getToken(ivFile, token); sscanf(token, "%f", &sr); getToken(ivFile, token); sscanf(token, "%f", &sg); getToken(ivFile, token); sscanf(token, "%f", &sb); tokenType = getToken(ivFile, token); } else if (SAME(token, "emissiveColor")) { getToken(ivFile, token); sscanf(token, "%f", &er); getToken(ivFile, token); sscanf(token, "%f", &eg); getToken(ivFile, token); sscanf(token, "%f", &eb); tokenType = getToken(ivFile, token); } else if (SAME(token, "shininess")) { getToken(ivFile, token); sscanf(token, "%f", &shin); tokenType = getToken(ivFile, token); } else if (SAME(token, verToken)) { for (vCount = 0; (tokenType=getToken(ivFile, token)) == T_NUMBER; vCount++) { if (vCount >= MAX_VERTICES) vCount = MAX_VERTICES - 1; sscanf(token, "%f", &v[vCount][0]); getToken(ivFile, token); sscanf(token, "%f", &v[vCount][1]); getToken(ivFile, token); sscanf(token, "%f", &v[vCount][2]); v[vCount][0] *= sx; v[vCount][1] *= sy; v[vCount][2] *= sz; } numV += vCount; strcpy(verToken, "voint"); } else if (SAME(token, "vector")) { for (nCount = 0; (tokenType=getToken(ivFile, token)) == T_NUMBER; nCount++) { if (nCount >= MAX_NORMALS) nCount = MAX_NORMALS - 1; sscanf(token, "%f", &n[nCount][0]); getToken(ivFile, token); sscanf(token, "%f", &n[nCount][1]); getToken(ivFile, token); sscanf(token, "%f", &n[nCount][2]); } numN += nCount; } else if (SAME(token, "filename")) { getToken(ivFile, token); sscanf(token, "%s", texPath); fprintf(stderr, "Texture Name: %s\n", texPath); pfLoadTexFile(style.texture, texPath); } else if (SAME(token, "point")) { for (tCount = 0; (tokenType=getToken(ivFile, token)) == T_NUMBER; tCount++) { if (tCount >= MAX_TEXTURES) tCount = MAX_TEXTURES - 1; sscanf(token, "%f", &t[tCount][0]); getToken(ivFile, token); sscanf(token, "%f", &t[tCount][1]); } numT += tCount; fprintf(stderr, "Texture Coordinates: %d\n", tCount); } else if (SAME(token, "numVertices")) { for (fCount = 0; (tokenType=getToken(ivFile, token)) == T_NUMBER; fCount++) { if (fCount >= MAX_INDICES) fCount = MAX_INDICES - 1; sscanf(token, "%d", &f[fCount]); } for (i = 0, j = 0; i < fCount; i++) { for (polygon.numVerts = 0; polygon.numVerts < f[i]; polygon.numVerts++) { pfCopyVec3(polygon.coords[polygon.numVerts], v[j]); if (nCount >= vCount) pfCopyVec3(polygon.norms [polygon.numVerts], n[j]); if (tCount >= vCount) pfCopyVec2(polygon.texCoords [polygon.numVerts], t[j]); j++; } if (polygon.numVerts > 2) { polygon.cbind = PFGS_PER_PRIM; polygon.nbind = (nCount >= vCount) ? PFGS_PER_VERTEX : PFGS_OFF; polygon.tbind = (tCount >= vCount) ? PFGS_PER_VERTEX : PFGS_OFF; numTris += polygon.numVerts - 2; pfuAddPoly(builder, &polygon); polygon.numVerts = 0; } } nCount = 0; vCount = 0; tCount = 0; pending = 0; } else if (tokenType == T_COMPLETE && pending) { if (viCount > 2) { for (i = 0, polygon.numVerts = 0; i < viCount; i++) { if (vi[i] >= 0 && vi[i] < MAX_INDICES) { pfCopyVec3(polygon.coords[polygon.numVerts], v[vi[i]]); if (nCount == vCount) pfCopyVec3(polygon.norms [polygon.numVerts], n[vi[i]]); polygon.numVerts++; } else if (polygon.numVerts > 2) { polygon.cbind = PFGS_PER_PRIM; polygon.nbind = (nCount == vCount) ? PFGS_PER_VERTEX : PFGS_OFF; polygon.tbind = PFGS_OFF; numTris += polygon.numVerts - 2; pfuAddPoly(builder, &polygon); polygon.numVerts = 0; } } } nCount = 0; vCount = 0; pending = 0; } else { /* * fprintf(stderr, " unrecognized: %s\n", token); */ tokenType = getToken(ivFile, token); ++numOther; } } /* close ".iv" file */ fclose(ivFile); /* construct geode */ if (numTris > 0) { geode = pfNewGeode(); node = (pfNode *)geode; processPolygons(geode, builder, &style); } /* delete builder */ pfuDelBuilder(builder); /* print statistics */ if(pfGetNotifyLevel() >= PFNFY_INFO) { elapsedTime = pfGetTime() - startTime; fprintf(stderr, "LoadIv:\n"); fprintf(stderr, " file name = %s\n", fileName); fprintf(stderr, " commands skipped = %8ld\n", numSkip); fprintf(stderr, " commands unknown = %8ld\n", numOther); fprintf(stderr, " vertices (indices) = %8ld (%8ld)\n", numV, numVI); fprintf(stderr, " normals (indices) = %8ld (%8ld)\n", numN, numNI); fprintf(stderr, " textures (indices) = %8ld (%8ld)\n", numT, numTI); fprintf(stderr, " triangles = %8ld\n", numTris); if (elapsedTime > 0.0) { fprintf(stderr, " loading time = %12.3f sec\n", elapsedTime); fprintf(stderr, " loading rate = %12.3f tri/sec\n", numTris/elapsedTime); } } /* print statistics */ if (pfGetNotifyLevel() >= PFNFY_INFO) fprintf(stderr, "\n"); return node; } static FILE * openFile (const char *fileName) { FILE *file = NULL; char filePath[MAX_STRING]; /* check argument */ if (fileName == NULL || *fileName == '\0') return NULL; /* find file in IRIS Performer directory-search path */ if (!pfFindFile(fileName, filePath, R_OK)) { pfNotify(PFNFY_WARN, PFNFY_RESOURCE, "openFile: Could not find file \"%s\"", fileName); return NULL; } /* open file */ if ((file = fopen(filePath, "r")) == NULL) { pfNotify(PFNFY_WARN, PFNFY_RESOURCE, "openFile: Could not open file \"%s\"", filePath); return NULL; } return file; } /* * make a reasonable default geostate */ static pfGeoState* defaultGeoState (void) { static void *arena = NULL; static pfMaterial *material = NULL; static pfGeoState *geostate = NULL; if (geostate == NULL) { arena = pfGetSharedArena(); material = pfNewMtl(arena); pfMtlColor(material, PFMTL_DIFFUSE, 0.5, 0.5, 0.5); pfMtlColor(material, PFMTL_AMBIENT, 0.5, 0.5, 0.5); pfMtlColorMode(material, PFMTL_FRONT, LMC_AD); geostate = pfNewGState(arena); pfGStateAttr(geostate, PFSTATE_FRONTMTL, material); } return geostate; } /* * add builder polygons to a geode */ static void processPolygons (pfGeode *geode, pfuBuilder *builder, Style *style) { int i; int numGeoSets = 0; pfList *geoSetList = NULL; pfGeoSet *geoset = NULL; pfGeoState *specific = NULL; pfMaterial *material = NULL; static void *sharedArena = NULL; if (geode == NULL || builder == NULL || style == NULL) return; if (sharedArena == NULL) sharedArena = pfGetSharedArena(); /* convert current face list into one or more geosets */ geoSetList = pfuMakeGSets(builder); numGeoSets = pfGetNum(geoSetList); /* make a specific geostate for this collection of polygons */ if (style->geostate != NULL) { specific = pfNewGState(sharedArena); pfCopy(specific, style->geostate); } /* set face culling mode to other-than-default */ if (style->culling == PFCF_OFF) pfGStateMode(specific, PFSTATE_CULLFACE, style->culling); /* set material to other-than-default */ material = pfNewMtl(sharedArena); pfMtlColor(material, PFMTL_AMBIENT, ar, ag, ab); pfMtlColor(material, PFMTL_DIFFUSE, dr, dg, db); pfMtlColor(material, PFMTL_SPECULAR, sr, sg, sb); pfMtlColor(material, PFMTL_EMISSION, er, eg, eb); pfMtlShininess(material, shin * 128.); pfMtlColorMode(material, PFMTL_FRONT, LMC_AD); pfGStateAttr(specific, PFSTATE_FRONTMTL, material); /* set texture to other-than-default */ if (style->texture != NULL) { pfGStateAttr(specific, PFSTATE_TEXTURE, style->texture); pfGStateAttr(specific, PFSTATE_TEXENV, style->environment); } /* bind specific geostate to each geoset and each geoset to the geode */ for (i = 0; i < numGeoSets; i++) { geoset = (pfGeoSet*)pfGet(geoSetList, i); pfGSetGState(geoset, specific); pfAddGSet(geode, geoset); } } /* * get next token from input */ static int getToken (FILE *fp, char *token) { int i; int c; /* find next token */ for (;;) { /* comsume space characters */ while ((c = getc(fp)) != EOF && isspace(c)) /* EMPTY */ {}; /* stop if end of file */ if (c == EOF) return T_END; /* consume comments */ if (c == '#' || c == '/') { while ((c = getc(fp)) != EOF && c != '\n') /* EMPTY */ {}; continue; } /* consume quoted strings */ if (c == '"') { i= 0; while ((c = getc(fp)) != EOF && c != '"') { token[i]= c; i++; }; token[i]= '\0'; return T_KEYWORD; } /* skip important grouping tokens */ if (c == '{' || c == ',' || c == '[' || c == ']' || c == '(' || c == ')' || c == '|' || c == '~' || c == '_' ) continue; /* extract keyword token */ if (c == '}') { token[0] = c; token[1] = '\0'; return T_COMPLETE; } if (c != EOF && isalpha(c)) { for (i = 0; c != EOF && (isalnum(c) || c == '_'); i++) { token[i] = c; c = getc(fp); } token[i] = '\0'; ungetc(c, fp); return T_KEYWORD; } else if (c != EOF && isdigit(c)||c=='.'||c=='+'||c=='-'||c=='e'||c=='E') { for (i=0; c!=EOF && (isdigit(c)||c=='.'||c=='+'||c=='-'||c=='e'||c=='E'); i++) { token[i] = c; c = getc(fp); } token[i] = '\0'; ungetc(c, fp); return T_NUMBER; } else fprintf(stderr, "<< %c >>\n", c); } } ... Cedric Cedric from Dahouet harbor on Armor coast in France - email: gce@syseca.fr ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ From guest Tue Aug 23 08:02:09 1994 Date: Tue, 23 Aug 1994 11:02:50 -0400 (EDT) From: "Lance R. Marrou" Subject: Re: Associating node callbacks with pfClone To: Kathryn Loynes Cc: Performer In-Reply-To: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Tue, 23 Aug 1994, Kathryn Loynes wrote: ... > I was wondering if it is possible to set up a node callback function > which would be invoked during a pfClone traversal. ... Good point. I bet many people would like to do this. It may not really be a large concern to traverse through a small structure of a model, but for a thousand models (coming and going) this could be a significant delay. _______________________________________________________ IST __ E-mail: marrou@vsl.ist.ucf.edu Visual / / ______ /\____ ______ ______ Systems / / / _ / / __ // ____// ____/ Lab / /__ / /_/ / / / / // /___ / __/_ R. Marrou ________/____//____/\\/_/ /_//_____//_____/____________ From guest Tue Aug 23 08:06:15 1994 (5.65b/IDA-1.4.3 for info-performer@sgi.com); Tue, 23 Aug 94 13:08:04 GMT From: Simon Gibbs Subject: video textures To: info-performer@sgi.sgi.com Date: Tue, 23 Aug 94 17:08:03 MDT We are interested in using Performer and a Sirius board to create video textures. The video input is a CCIR 601 4:4:4:4 signal. We want to create a 4 component texture and then use subtexload to dynamically modify the texture. The problem: subtexload needs a texture id and these are not visible when using Performer. Is there a Performer equivalent of subtexload or a way of getting hold of texture ids for textures created with pfNewTex? ---------------------------------------------------------------------------- Simon Gibbs GMD Schloss Birlinghoven 53574 Sankt Augustin GERMANY ---------------------------------------------------------------------------- Email: Simon.Gibbs@gmd.de WWW: http://cui_www.unige.ch/OSG/Simon Tel: +49 (2241) 14.2697 (direct) +49 (2241) 14.2366 (department office) +49 (2241) 14.2040 (fax) ---------------------------------------------------------------------------- From guest Tue Aug 23 12:56:57 1994 Date: Tue, 23 Aug 1994 14:56:01 -0500 (CDT) From: Olukayode B Olofinboba Subject: Re: John Burwell or IRIS performer SGI contact To: Norbert BIANCHIN Cc: info-performer@sgi.sgi.com In-Reply-To: <9408230841.ZM26756@shark.paris.sgi.com> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Thanks Norbert. - Olu From guest Tue Aug 23 13:35:24 1994 Date: Tue, 23 Aug 1994 13:33:41 -0700 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: info-performer@sgi.sgi.com From: jan@archimedes.chinalake.navy.mil (Jan A. Barglowski) Subject: Sectioning a texture? Hello all! I've built a rudimentary terrain builder, and I'm texturing it with a scanned-in image of a map. The image file (rgb) is in one piece, but I've implemented a terrain sectioning algorithm that breaks down the geometry recursively and spatially, eg. 1 section into 4, then each of the 4 breaks down, etc. I can map the texture vertices correctly on the sub-sections, but the texture file is getting read for each piece, and fills up texture memory with a lot of redundant imagery! I've already tried making one big terrain, then using pfuBreakup on it, but I need to have different LOD's. Question is, how do I read only one copy of the texture, then split it up somehow to cover the individual sub-sections? Only was I can figure so far is to perform my own "Breakup" of the terrain. Any help or ideas appreciated! jan. Jan Anthony Barglowski Phone: (619) 927-1057 Computer Dude Internet: jan@archimedes.nwc.navy.mil Visualization Lab Packet: SYSOP@WA6YBN.#SOCA.CA.US.NA Naval Air Warfare Center kc6uth@kc6uth.ampr.org -Weapons Division [44.17.2.6] From guest Tue Aug 23 17:55:12 1994 To: gce@scl.syseca.fr (Cedric Gautier ) Cc: info-performer@sgi.sgi.com Subject: Re: Performer with ImageVision, Movie and Video libraries ... In-Reply-To: Your message of "Tue, 23 Aug 1994 11:39:22 +0200." <9408230939.AA21444@scl> Date: Wed, 24 Aug 1994 09:12:29 +0900 From: Ryutarou OHBUCHI You have mentioned what sounded like a live video pasted on polygons (as texture) using Performer on an Indy. Can I do that on an Onyx with RE2? ----- Ryutarou Ohbuchi IBM Japan, Ltd., SA-S76 1623-14, Shimo-tsuruma, Yamato-shi, Kanagawa-ken 242, Japan Phone: 0462-73-4882 Fax: 0462-73-7413 ip: ohbuchi@trl.ibm.co.jp From guest Wed Aug 24 14:01:03 1994 Date: Wed, 24 Aug 94 10:33:44 -0700 From: lessin@indy2.gstone.com (Dan Lessin) To: info-performer@sgi.sgi.com Subject: Contouring Hi. There is a demo on the Onyx where I work called 'Contouring' and I was wondering if this was some kind of encapsulated feature, or is it just a nice effect coded up by sgi's software guys ( In other words, how can I do it ). The effect involves a three-dimensional texture ( or maybe just a projection of a 2D texture ) which remains stationary as an object moves through it. The stationary texture mapped onto the moving object creates a unique effect. I don't know too much about the details of texture mapping, and would appreciate any ideas anyone might have on the best way to recreate this. Thanks. -- Dan Dan Lessin Humble Intern Greystone Technology lessin@realityone.gstone.com ( I think ) From guest Wed Aug 24 14:02:39 1994 Date: Wed, 24 Aug 1994 09:41:01 -0700 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: info-performer@sgi.sgi.com From: jan@archimedes.chinalake.navy.mil (Jan A. Barglowski) Subject: Sectioning textures: figured out... Sorry, turned out to be a dumb technique I used which assigned a new GeoState to each terrain section. All I had to do was create it once and then "share" it! jan. Jan Anthony Barglowski Phone: (619) 927-1057 Computer Dude Internet: jan@archimedes.nwc.navy.mil Visualization Lab Packet: SYSOP@WA6YBN.#SOCA.CA.US.NA Naval Air Warfare Center kc6uth@kc6uth.ampr.org -Weapons Division [44.17.2.6] From guest Wed Aug 24 14:05:20 1994 Date: Wed, 24 Aug 94 10:06:50 CDT From: lac@gothamcity.jsc.nasa.gov (Lac Nguyen) Apparently-To: Hi every one, Would some body some me how to do stereo using the Crystal Eyes using Performer. A way back some body have make the perfly.c work with Crystal Eyes, and put it out in public. But I don't know where ftp site. Did any one know ? Thank a lot. Lac From guest Wed Aug 24 14:06:44 1994 From: bleclerc@cae.ca (Bernard Leclerc) Subject: Dynamically loading (paging) a large database To: info-performer@sgi.sgi.com Date: Wed, 24 Aug 1994 09:20:19 -0400 (EDT) X-Mailer: ELM [version 2.4 PL21] Content-Type: text Content-Length: 858 Hi Performer fans, I'm fairly new to Performer but I've been in the real time simulation domain for a while. After reading part I of the IRIS Performer Programming Guide, I have full of questions to ask. Here is the first one. Our application requires very large databases - one covers 1100 km by 900 km. It's clearly not possible to load the entire database in memory at init time. Our approach consist in having a separate loader whose task is to monitor the viewpoint and to load the surrounding tiles. At every iteration, the whole Performer hierarchy is rebuilt from the loaded tiles. Is there a better way of approaching the problem? How can the Performer hierarchy be updated at run time with new terrain data and how can we remove old data from it? -- Bernard Leclerc CAE Electronics Ltd. tel.: +1 514 341 2000, local 2275 e-mail: bleclerc@cae.ca From guest Wed Aug 24 14:07:36 1994 Date: Wed, 24 Aug 1994 10:49:47 GMT From: cseal@sirapgc.demon.co.uk (Chris Seal) Reply-To: cseal@sirapgc.demon.co.uk To: info-performer@sgi.sgi.com Subject: Performer with CrystalEyes VR glasses X-Mailer: PCElm 1.10 Lines: 15 Hi, I'm new to info-performer so please forgive me if this has been previously covered. I'm intending to use Performer with StereoGraphics CrystalEyes VR glasses. Has anyone developed an interface for these glasses to control the viewpoint in Performer (or for Logitech's 3D mouse). I would also be interested to know of any existing applications which use the Performer and CrystalEyes VR glasses Many thanks -- Chris Seal From guest Wed Aug 24 18:54:29 1994 From: davec@cory.coryphaeus.com (David Cooper) Subject: Re: Dynamically loading (paging) a large database To: info-performer@sgi.sgi.com Date: Wed, 24 Aug 1994 17:40:22 -0800 (PDT) In-Reply-To: <9408241320.AA16282@cae.ca> from "Bernard Leclerc" at Aug 24, 94 09:20:19 am X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 645 > approaching the problem? How can the Performer hierarchy be updated at run time > with new terrain data and how can we remove old data from it? I'd be very interested in an answer to this, particularly ways around the current limitation of a user spawned thread only being able to create pf(things) when the APP is not in pfSync or pfFrame. As far as I can see this limitation makes it impossible to use another process to do terrain paging for you. In effect, every pf(create anything) must be in the app. In my case I have at least one cpu doing virtually nothing, and the app processor overloaded, and no way to share. DaveC From guest Thu Aug 25 04:22:15 1994 From: "JAVIER CASTELLAR" Date: Thu, 25 Aug 1994 23:22:19 -0600 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: ohbuchi@trl.ibm.co.jp Subject: Re:Re: Performer with ImageVision, Movie and Video libraries ... Cc: info-performer@sgi.sgi.com, javier@madrid.sgi.com Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 Hi, Yes, you can do it. There are two ways to texture live video onto any kind of geometry (true real time video). a) Using an OPTIONAL piece of hardware: Sirius Video. a.1) It can give you a direct path from CIR601 sources up to texture memory. Mainly it can update textures from video. In order to use this feature under performer you must use a couple of GL callbacks to use the full functionality of texdef2d (FAST_DEFINE feature), subtexload (null input ) as well as a couple of VL (Video Library) calls to stablish the path from the desired video source, to fix the format (4:2:2, 4:4:4, 4:4:4:4 , RGB, RGBA ) and the drain onto texture memory. a.2) It can give you a direct path from CIR601 sources up to main memory. Mainly it can update textures from memory. In order to use this feature under performer you must use a couple of GL callbacks to use the full functionality of texdef2d (FAST_DEFINE feature), subtexload (null input ) as well as a couple of VL (Video Library) calls to stablish the path from the desired video source, to fix the format (4:2:2, 4:4:4, 4:4:4:4 , RGB, RGBA, YUV, YUVY ) and the drain onto memory using the VME bus.Then from main memory you can use this info to define textures us fields. a.3) Use the path from video to memory to load at the same time in hard disk (you will need a highly stripped hard disk system, remember that our mahines can support docens of SCSI2 F/W 20MB/s). Then from disk you can create your own processes to reload it like another video sources. b) Creating your own movies in memory and use it as above, but using the proper subtexload parameters on the PreDraw Callback. Just remember all the real time issues related with this kind of highly couple parallel processes. You will need in some cases spawn you own processes in another CPUs sharing data with Performer APP & ISECT processes. People use to work in this way very well at funcionality level, i mean that it works. You share data using the same arenas that you create in Performer. You can use fork() and sproc() at starting stages to create your own user processes. In order to sustain frame rate levels (25Hz or 30Hz), or even field rates (50Hz or 60Hz), you will need to make up a good real time design (semaphores properly allocated, proper barriers, ... proper design). You will need also to tune some kernel parameters for take a full use of main memory (there is a couple of variable to let a process to use all the memory), as well us disabling interrupts on DRAW and CULL cpus, as well us your own procesess if needed or suitable in function of the hardware used (VME devices, Sirius ... etc), i am talking about the NOINTR: directives and so on.[Please refer to the REACT technical reports] Hope to help. Javier Castellar Systems Engineer javier@madrid.sgi.com From guest Thu Aug 25 07:47:41 1994 Thu, 25 Aug 94 16:46:52 GMT From: "Aris Christidis" Organization: AITEC GmbH & Co KG To: info-performer@sgi.sgi.com Date: Thu, 25 Aug 1994 16:46:40 GMT+1 Subject: gamma correction Cc: kp@aitec.de, michaels@cologne.sgi.com Priority: normal X-Mailer: Pegasus Mail v3.1 (R1a) I need a way to change the display gamma at Multi-Channel Option. "vout" is no solution, as it changes the gamma slope but not the function itself, and there are at least two default settings, the exponential (obviously used for general purposes - see GL Programming Guide, p.4-24) and the linear one (problably applied, when MCO is run - see on-line man pages). The use of "gammaramp(r,g,b)" helps manipulate the gamma in an own application program, but it has no effect, when MCO is on. Does anybody know a solution to this problem? AITEC GmbH & CO Informationstechnologie KG Dr. Aris Christidis Alter Hellweg 50 D-44379 Dortmund Tel.: +49 231 9646546 Fax.: +49 231 9646598 EMail: ac@aitec.de From guest Thu Aug 25 10:55:14 1994 Date: Thu, 25 Aug 1994 10:55:22 -0700 From: hitchner@netcom.com (Lew Hitchner) To: AC@AITEC.de, info-performer@sgi.sgi.com Subject: Re: gamma correction Dr. Aris Christidis from Dortmund, Germany writes: > I need a way to change the display gamma at Multi-Channel Option. > "vout" is no solution, as it changes the gamma slope but not the > function itself, and there are at least two default settings, the >> exponential (obviously used for general purposes - see GL Programming > Guide, p.4-24) and the linear one (problably applied, when MCO is run > - see on-line man pages). > The use of "gammaramp(r,g,b)" helps manipulate the gamma in an own > application program, but it has no effect, when MCO is on. > Does anybody know a solution to this problem? I cannot answer his question directly, but have some suggestions. Perhaps an SGI employee will give him a complete answer to his question. When in a previous job I was using a Video Splitter (a four video channel predecessor of the MCO) on our 4D/440 IG2 SkyWriter under IRIX 4.0.5, I ran into bugs in the software for gamma correction. The vs software supposedly supported gamma correction that allows a user application to directly load the 256 entry r,g,b compensation table values with any function desired. However, when I tried using that fn., it didn't work at all (I think it just set the table to all 1's or all 0's or something dumb like that). We had full customer support, so I logged a call. I was assigned to a specialist in video software support. He told me, yes the fn. didn't work and that it was fixed in the next release of the VideoSplitter software which was to come out with the next release of IRIX. I left my old job before that happened, so I don't know what's happened. If you have a support contract, here's what you might try. Place a call and get a call # and give SGI your email address. I don't know how making the phone call is best accomplished from Europe. But once you get he call # and a name of the support person assigned, you should be able to do all you communication via email thereafter. Thus, you could minimize the pain of time zone differences between Calif. and Germany. That's what I did when I called in my VideoSplitter support request, even tho I was at NASA Ames only about a mile from SGI HQ. If you don't have full support, then hope that someone in info-performer can help you. Lew Hitchner Virtual Reality and Visual Simulation Consultant Mountain View, CA From guest Thu Aug 25 14:53:22 1994 To: Kathryn Loynes Cc: Performer Subject: Re: Associating node callbacks with pfClone In-Reply-To: Your message of "Tue, 23 Aug 94 18:04:36 +1000." Date: Thu, 25 Aug 94 14:53:03 -0700 From: Jim Helman > I can't see any obvious way of doing this, if anyone has any ideas I'd > be interested to hear them. Your approach seems reasonable. The inability of Performer applications to control or gain visibility into the pfClone traversal is a deficiency which we'll try to address in the next release. rgds, -jim helman jimh@surreal.asd.sgi.com 415/390-1151 From guest Thu Aug 25 15:04:53 1994 To: lac@gothamcity.jsc.nasa.gov (Lac Nguyen) Cc: info-performer@sgi.sgi.com Subject: Crystal Eyes with Performer In-Reply-To: Your message of "Wed, 24 Aug 94 10:06:50 CDT." <9408241506.AA27971@gothamcity.jsc.nasa.gov> Date: Thu, 25 Aug 94 15:04:33 -0700 From: Jim Helman > A way back some body have make the perfly.c work with Crystal Eyes, and > put it out in public. But I don't know where ftp site. > Did any one know ? sfly.tar.Z, which is a stereo-in-a-window version of Perfly, can be found on sgigate.sgi.com:/pub/Performer. We don't have a version that supports head-tracking (CrystalEyes VR), but hope to put some code out this fall. Unfortunately, someone destroyed the LCDs in my glasses, so I may only get to see head-tracked mono ;-{. rgds, -jim helman jimh@surreal.asd.sgi.com 415/390-1151 From guest Fri Aug 26 15:00:59 1994 From: "Drew Hess" Date: Fri, 26 Aug 1994 10:59:32 -0700 X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: iris-performer@sgi.sgi.com Subject: Identifying channels in callbacks Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 I need a way of uniquely identifying the channel passed to a post/predraw callback in Performer 1.2. According to the man page, the return value of pfNewChan() is a handle to the newly created channel, so it sounds like I could use the handle for what I need. (What's actually returned is of type (pfChannel *), so should I use the pointer itself to identify the channel or should I dereference the pointer and use that value? The latter would work if, for instance, the pointer points to an integer that's used as a table index into a libpr or libpf internal table.) Is this a valid assumption? Are these handles static throughout the run-time life of the program? (i.e. can I assume that Performer will not go behind my back and decide to change the value of the handle sometime while the program is running?) Do they have the same value when passed to different processors on an MP machine running Performer with separate app, cull, and draw processes? Thanks for any answers, -dwh- dhess@vision.arc.nasa.gov From guest Fri Aug 26 15:16:35 1994 Date: Fri, 26 Aug 94 11:51:38 EDT From: louis@speech.iar.nrc.ca (Louis Bolduc) To: info-performer@sgi.sgi.com Subject: pfExit() question... Hi everybody! You are all familiar with the documented bug in 'pfExit()' which causes the whole application to terminate when it is called. Suppose I want to terminate the application, but I want to return a status code different than 0 (which is what is returned when you invoke 'pfExit()'). Will a call to 'exit(status)' instead of pfExit() be sufficient? will there be some resources allocated by Performer left around, which should be cleaned up? - Louis | Life would be much easier to| Louis Bolduc | === __ | | understand if mother nature | (613) 998-9780 | \_______| _) | | gave us the source code. | louis@speech.iar.nrc.ca | __ /\ | __ | | - Graeme MacWilliam | Flight Research | / / \ / \ \ | | | Laboratory | | /_|_\/\/ | \ | | | #include | | \ / \/ \ / | +----- NRC/CNRC - Institute for Aerospace Research -----+- \__/ ------- \__/ -+ From guest Fri Aug 26 15:22:16 1994 Date: Fri, 26 Aug 94 16:14:30 +0900 From: wry@dimwit.dst.nk-exa.co.jp (Masahiko Yamanaka) To: info-performer@sgi.sgi.com Subject: Stereo on HiVision I want to do StereoView at HDTV Resolution( Actually, 1920x about500 ). So, this needs following 2 devices. 1) HDTV which supports (1920x500@120Hz)or(1920x500@100Hz) as the input. 2) Graphics Computer which supports such an output signal. But, I've never heared of these devices. Have anyone heard of such a 1) device? if 1) exists, will SGI Products support such a signal as the Output? Any information will be appreciated. -- Masahiko Yamanaka From jimh@surreal Fri Aug 26 21:03:51 1994 To: wry@dimwit.dst.nk-exa.co.jp (Masahiko Yamanaka) Cc: info-performer@sgi.sgi.com Subject: Re: Stereo on HiVision In-Reply-To: Your message of "Fri, 26 Aug 94 16:14:30 +0900." <9408260714.AA07671@dimwit.dst.nk-exa.co.jp> Date: Fri, 26 Aug 94 19:00:06 -0700 From: Jim Helman > 1) HDTV which supports (1920x500@120Hz)or(1920x500@100Hz) as the input. I don't know of any HDTV monitors that can handle stereo. Also, the bandwidth exceeds what RealityEngine can produce, since our hidef video formats are 1600x1200@60Hz and 1920x1035@30Hz-interlaced. With a suitable monitor and a custom RealityEngine video format (or an external converter), something like 1600x600@120Hz or 1920x500@60Hz might be possible, at least technically. In sort, I don't think there's an off-the-shelf way to do put a system together to do stereo in this way. Another approach would be to use two hidef video projectors with polarization filters, each driven by a RealityEngine graphics pipeline. This has the advantage that you'd get the full vertical resolution, could use regular polarizing glasses rather than shutter glasses, would have more graphics processing power, and can be assembled from commercially available components. rgds, -jim helman jimh@surreal.asd.sgi.com 415/390-1151 From jimh@surreal Fri Aug 26 21:03:17 1994 To: zrfu0390@awssg7.rus.uni-stuttgart.de (Dirk Rantzau) Cc: info-performer@sgi.sgi.com Subject: Re: Crystal Eyes with Performer In-Reply-To: Your message of "Fri, 26 Aug 94 13:15:32 MDT." <9408261115.AA29435@awssg7.rus.uni-stuttgart.de> Date: Fri, 26 Aug 94 18:36:01 -0700 From: Jim Helman sfly uses stereo-in-a-window (leftbuffer/rightbuffer) which allows stereo and non-stereo apps to coexist on the screen. Unfortunately, so far it's only available on RealityEngine, where you can use /usr/gfx/setmon to set a stereo video format, e.g. setmon -n 1025x768_96s. Porting sfly to the old-style STR_RECT isn't difficult. Just remove the leftbuffer/rightbuffer calls and set up the viewports of the pfChannels (pfChanViewPort) to occupy the correct portions of a full screen pfPipe. rgds, -jim helman jimh@surreal.asd.sgi.com 415/390-1151