From owner-info-inventor-dev@oss.sgi.com Fri Dec 1 11:07:01 2000 Received: by oss.sgi.com id ; Fri, 1 Dec 2000 11:06:51 -0800 Received: from deliverator.sgi.com ([204.94.214.10]:57918 "EHLO deliverator.sgi.com") by oss.sgi.com with ESMTP id ; Fri, 1 Dec 2000 11:06:44 -0800 Received: from kinabalu.csd.sgi.com (kinabalu.csd.sgi.com [150.166.1.88]) by deliverator.sgi.com (980309.SGI.8.8.8-aspam-6.2/980310.SGI-aspam) via ESMTP id LAA09352 for ; Fri, 1 Dec 2000 11:06:43 -0800 (PST) mail_from (jlim@kinabalu.csd.sgi.com) Received: by kinabalu.csd.sgi.com (980427.SGI.8.8.8/911001.SGI) id LAA19015; Fri, 1 Dec 2000 11:04:33 -0800 (PST) From: jlim@kinabalu.csd.sgi.com (Jonathan Lim) Message-Id: <200012011904.LAA19015@kinabalu.csd.sgi.com> Subject: Re: GLwMDrawA To: walsh_tj@mail.one.net.au Date: Fri, 1 Dec 2000 11:04:33 -0800 (PST) Cc: info-inventor-dev@oss.sgi.com X-Mailer: ELM [version 2.5 PL2] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing Forwarded message: > > Guys, > > I have had some problems compiling on my Mandrake 7.2 system. The only > thing I've changed since installing Mandrake is the XFree86 4.0.1 driver to > the nvidia binary driver, including changing the Mesa libraries to nVidia's > - I don't know if that could have caused the problem? Anyway, attached is > the output of my call to sudo gmake. The file that the build looks for and > can't find (GL/GLwMDrawA.h) is one that I have never heard of on any Linux > system I've used. Could it be a SGI file? I can't find the file under any It's not exclusive to SGI. The Mesa distribution has it under Mesa-3.2.1/widgets-sgi, so you could build it yourself easily. > packages that the build is supposed to depend on, including Mesa/nVidia, > Lesstif, XFree86 4.0.1. I've even looked through file lists of older > versions of these packages on http://rufus.w3.org. I will be using > OpenInventor for my engineering honours research project next year on SGI > hardware but I'd love to be able to develop at home as well. > > Thanks for your help, > Toby Jonathan Lim _ Silicon Graphics _ Mountain View GPS Graphics Computer Systems CA, USA From owner-info-inventor-dev@oss.sgi.com Fri Dec 1 15:18:53 2000 Received: by oss.sgi.com id ; Fri, 1 Dec 2000 15:18:43 -0800 Received: from jalon.able.es ([212.97.163.2]:40157 "EHLO jalon.able.es") by oss.sgi.com with ESMTP id ; Fri, 1 Dec 2000 15:18:15 -0800 Received: from correo.able.es ([212.97.169.121]) by jalon.able.es (Netscape Messaging Server 4.15) with SMTP id G4WWWK00.84O; Sat, 2 Dec 2000 00:21:56 +0100 Date: Sat, 2 Dec 2000 00:18:02 +0100 From: "J . A . Magallon" To: Jonathan Lim Cc: walsh_tj@mail.one.net.au, info-inventor-dev@oss.sgi.com Subject: Re: GLwMDrawA Message-ID: <20001202001802.A3982@werewolf.able.es> Reply-To: jamagallon@able.es References: <200012011904.LAA19015@kinabalu.csd.sgi.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In-Reply-To: <200012011904.LAA19015@kinabalu.csd.sgi.com>; from jlim@kinabalu.csd.sgi.com on Fri, Dec 01, 2000 at 20:04:33 +0100 X-Mailer: Balsa 1.0.0 Content-Length: 1643 Lines: 34 Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing On Fri, 01 Dec 2000 20:04:33 Jonathan Lim wrote: > Forwarded message: > > > > Guys, > > > > I have had some problems compiling on my Mandrake 7.2 system. The only > > thing I've changed since installing Mandrake is the XFree86 4.0.1 driver to > > the nvidia binary driver, including changing the Mesa libraries to nVidia's > > - I don't know if that could have caused the problem? Anyway, attached is > > the output of my call to sudo gmake. The file that the build looks for and > > can't find (GL/GLwMDrawA.h) is one that I have never heard of on any Linux > > system I've used. Could it be a SGI file? I can't find the file under any > > It's not exclusive to SGI. The Mesa distribution has it under > Mesa-3.2.1/widgets-sgi, so you could build it yourself easily. > The GLX widgets were added to XFree in version 4.0.1f. Now you have 4.0.1g in rufus, that also includes GLw and many more bug fixes. The 'bad' news are that I think it requires glibc2.2. But I have upgraded to glibc2.2 and all seems to work fine. You need a good session of rpm upgrading, but at the end all works fine. Now I am running Performer and Inventor on an up-to-date cooker system (glibc22, gcc2.96, XFree401g). Seems to be stable (last night I left town.perfly running for a couple hours...). Note: Mesa included in XFree is 3.4. That forces to make some changes to Inventor to compile, but that is subject for one other mail... -- Juan Antonio Magallon Lacarta #> cd /pub mailto:jamagallon@able.es #> more beer Linux 2.2.18-pre24-vm #2 SMP Wed Nov 29 02:56:21 CET 2000 i686 unknown From owner-info-inventor-dev@oss.sgi.com Fri Dec 1 15:49:53 2000 Received: by oss.sgi.com id ; Fri, 1 Dec 2000 15:49:34 -0800 Received: from jalon.able.es ([212.97.163.2]:13534 "EHLO jalon.able.es") by oss.sgi.com with ESMTP id ; Fri, 1 Dec 2000 15:49:25 -0800 Received: from correo.able.es ([212.97.169.121]) by jalon.able.es (Netscape Messaging Server 4.15) with SMTP id G4WYCJ00.661 for ; Sat, 2 Dec 2000 00:53:07 +0100 Date: Sat, 2 Dec 2000 00:49:14 +0100 From: "J . A . Magallon" To: OpenInventor Devel List Subject: [PATCH] Linux Makefiles Message-ID: <20001202004914.E3982@werewolf.able.es> Reply-To: jamagallon@able.es Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="AWniW0JNca5xppdA" Content-Transfer-Encoding: 8bit X-Mailer: Balsa 1.0.0 Content-Length: 6682 Lines: 208 Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing --AWniW0JNca5xppdA Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Hi everyone. I have been working to compile OpenInventor on a Linux Mandrake 7.2 - Cooker system. I have made some changes to Makefiles in order to make them more readable or suit better to the linux file system tree specs. Patch is attached. I think it can be included in the CVS tree. Changes affect make/ivcommonrules, make/commondefs and make/ivcommondefs. Changes: - Use 'make -Cdir' instead of '(cd $$i && $(MAKE))'. Less shell spawning. This can only be used with GNU make (IRIX make has not the -C flag), but as the README.FIRST file says it has to be built wi gmake... - If use make -Cdir, don't test if dir exists. If it is not there, make will complain and give the error msg that now is done manually. Also informs about the directory change ('entering....' and 'leaving...'). - Changed test-dir-and-make-it from if [ ! -d dir ]; then mkdir -p dir; fi to mkdir -p dir This also works on IRIX. If the dir exists, mkdir -p silently ignores, so no test needed. Also mkdir -p can handle multiple arguments. - bzip2'ing of man pages. - location of files. In linux there should not be any /usr/demos. All goes in /usr/share. Specific changes are: * man1 to /usr/share/man/man1 instead of /usr/man/man1 * man3 to /usr/share/man/man3 instead of /usr/man/man3 * help to /usr/share/Inventor/help instead of /usr/share/help * data to /usr/share/Inventor/data instead of /usr/share/data * demos to /usr/share/Inventor/demos instead of /usr/share/demos All prefixed with $(IVROOT) Hope you agree with the changes and introduce them in CVS. Please mail me or the list if you do to update my CVS tree. -- Juan Antonio Magallon Lacarta #> cd /pub mailto:jamagallon@able.es #> more beer Linux 2.2.18-pre24-vm #2 SMP Wed Nov 29 02:56:21 CET 2000 i686 unknown --AWniW0JNca5xppdA Content-Type: application/octet-stream; charset=us-ascii Content-Disposition: attachment; filename=patch-make --- ./inventor/make/ivcommonrules.org Sun Nov 19 01:45:31 2000 +++ ./inventor/make/ivcommonrules Fri Nov 24 03:20:41 2000 @@ -1,7 +1,5 @@ include $(COMMONRULES) - - # # Rule for generating C++ from shape source code in PP format. # @@ -25,21 +23,21 @@ $(LD) $(LDDSOOPTS) $^ $(LDLIBS) $(POST_LDDSOOPTS) -o $@ $(LARCHIVE): - for i in $(SUBDIRS); do (cd $$i && $(MAKE) all) || exit 1; done + for i in $(SUBDIRS); do $(MAKE) -C$$i all || exit 1; done $(RM) $(LARCHIVE) $(AR) $(ARFLAGS) $(LARCHIVE) $(OBJECTS) install_ivlibdso: $(LIBDSO) - if [ ! -d $(IVLIBDIR) ]; then mkdir -p $(IVLIBDIR); fi + mkdir -p $(IVLIBDIR) $(INSTALL) -m 755 $< $(IVLIBDIR) if [ "$(LIBTYPE)" = "debug" ]; then \ - if [ ! -d $(IVDLIBDIR) ]; then mkdir -p $(IVDLIBDIR); fi; \ + mkdir -p $(IVDLIBDIR); \ $(INSTALL) -m 755 $< $(IVDLIBDIR); \ fi install_ivheaders: - if [ ! -d $(IVHDRDIR) ]; then mkdir -p $(IVHDRDIR); fi + mkdir -p $(IVHDRDIR); for h in $(HEADERS); do $(INSTALL) -m 644 $$h $(IVHDRDIR); done LDIRT += $(LARCHIVE) @@ -60,12 +58,18 @@ all_man3iv: $(MAN3PAGES) install_man3iv: $(MAN3PAGES) - if [ ! -d $(IVMAN3DIR) ]; then mkdir -p $(IVMAN3DIR); fi - for i in $^; do $(INSTALL) -m 644 $$i $(IVMAN3DIR); done + mkdir -p $(IVMAN3DIR) + for i in $^; do \ + $(INSTALL) -m 644 $$i $(IVMAN3DIR); \ + bzip2 -f $(IVMAN3DIR)/$$i; \ + done install_man1: $(MAN1PAGES) - if [ ! -d $(IVMAN1DIR) ]; then mkdir -p $(IVMAN1DIR); fi - for i in $^; do $(INSTALL) -m 644 $$i $(IVMAN1DIR); done + mkdir -p $(IVMAN1DIR) + for i in $^; do \ + $(INSTALL) -m 644 $$i $(IVMAN1DIR); \ + bzip2 -f $(IVMAN1DIR)/$$i; \ + done LDIRT += $(MAN3PAGES) @@ -95,19 +99,16 @@ $(LD) $(LDDSOOPTS) $(OBJECTS) $(LDLIBS) $(POST_LDDSOOPTS) -o $@ install_ivbin: $(PROGRAM) - if [ ! -d $(IVBINDIR) ]; then mkdir -p $(IVBINDIR); fi + mkdir -p $(IVBINDIR) $(INSTALL) -m 755 $(PROGRAM) $(IVBINDIR) install_ivscript: - if [ ! -d $(IVBINDIR) ]; then mkdir -p $(IVBINDIR); fi + mkdir -p $(IVBINDIR) $(INSTALL) -m 755 $(SCRIPT) $(IVBINDIR) install_ivdemo: $(DEMO) - if [ ! -d $(IVDEMOBINDIR) ]; then mkdir -p $(IVDEMOBINDIR); fi - if [ ! -d $(IVDEMODATADIR) ]; then mkdir -p $(IVDEMODATADIR); fi + mkdir -p $(IVDEMOBINDIR) $(IVDEMODATADIR) $(INSTALL) -m 755 $(DEMO) $(IVDEMOBINDIR) - - # # Convenience assignments. TARGETS are removed by 'make clobber'. --- ./inventor/make/commondefs.org Sun Nov 19 01:41:58 2000 +++ ./inventor/make/commondefs Fri Nov 24 03:08:58 2000 @@ -663,14 +663,7 @@ # are not present. This is useful with mandefs/rules # SUBDIR_MAKERULE= \ - if test ! -d $$d; then \ - if test "$(NOSUBMESG)" = "" ; then \ - echo "SKIPPING $$d: No such directory."; \ - fi \ - else \ - echo " (cd $$d && $(MAKE) $${RULE:=$@})"; \ - (cd $$d && ${MAKE} $${RULE:=$@}) || exit 1; \ - fi + $(MAKE) -C$$d $${RULE:=$@} || exit 1 SUBDIRS_MAKERULE= \ @for d in $(SUBDIRS); do $(SUBDIR_MAKERULE); done @@ -695,14 +688,7 @@ # @$(NSUBDIR_MAKERULE) NSUBDIR_MAKERULE= \ - if test ! -d $@; then \ - if test "$(NOSUBMESG)" = "" ; then \ - echo "SKIPPING $@: No such directory."; \ - fi \ - else \ - echo " (cd $@; $(MAKE) $(__DORULE))"; \ - (cd $@; ${MAKE} $(__DORULE)); \ - fi + $(MAKE) -C$@ $(__DORULE) # # The macro naming commonrules' always-unsatisfied target, which is useful --- ./inventor/make/ivcommondefs.org Fri Nov 24 01:29:37 2000 +++ ./inventor/make/ivcommondefs Fri Nov 24 02:00:19 2000 @@ -64,9 +64,9 @@ # Determine optimization from LIBTYPE. # -OPTIMIZER = -O -DNDEBUG +OPTIMIZER = -O2 -march=i586 -DNDEBUG ifneq (, $(findstring debug, $(LIBTYPE))) -OPTIMIZER = -g +OPTIMIZER = -g -O2 -march=i586 endif @@ -105,16 +105,16 @@ IVBINDIR = $(IVROOT)/usr/bin IVLIBDIR = $(IVROOT)/usr/lib -IVDLIBDIR = $(IVLIBDIR)/InventorDebug +IVDLIBDIR = $(IVLIBDIR)/debug IVHDRTOP = $(IVROOT)/usr/include/Inventor IVHDRDIR = $(IVHDRTOP)/$(HDRSUBDIR) -IVMAN1DIR = $(IVROOT)/usr/man/man1 -IVMAN3DIR = $(IVROOT)/usr/man/man3 -IVHELPDIR = $(IVROOT)/usr/share/help -IVDATADIR = $(IVROOT)/usr/share/data/models -IVDEMODIR = $(IVROOT)/usr/demos -IVDEMOBINDIR = $(IVDEMODIR)/Inventor -IVDEMODATADIR = $(IVDEMODIR)/data/Inventor +IVMAN1DIR = $(IVROOT)/usr/share/man/man1 +IVMAN3DIR = $(IVROOT)/usr/share/man/man3 +IVHELPDIR = $(IVROOT)/usr/share/Inventor/help +IVDATADIR = $(IVROOT)/usr/share/Inventor/data +IVDEMODIR = $(IVROOT)/usr/share/Inventor/demos +IVDEMOBINDIR = $(IVDEMODIR) +IVDEMODATADIR = $(IVDEMODIR)/data --AWniW0JNca5xppdA-- From owner-info-inventor-dev@oss.sgi.com Fri Dec 1 16:09:23 2000 Received: by oss.sgi.com id ; Fri, 1 Dec 2000 16:09:04 -0800 Received: from mail.disney.com ([204.128.192.15]:45722 "EHLO mail.disney.com") by oss.sgi.com with ESMTP id ; Fri, 1 Dec 2000 16:08:41 -0800 Received: from pain10.corp.disney.com (root@pain10.corp.disney.com [153.7.110.100]) by mail.disney.com (Switch-2.0.1/Switch-2.0.1) with SMTP id eB208Zp26066 for ; Fri, 1 Dec 2000 16:08:35 -0800 (PST) Received: from louie.fa.disney.com by pain.corp.disney.com with ESMTP; Fri, 1 Dec 2000 16:09:19 -0800 Received: from plio.fan.fa.disney.com (plio.fan.fa.disney.com [153.7.118.2]) by louie.fa.disney.com (8.9.2/8.9.2) with ESMTP id QAA11193; Fri, 1 Dec 2000 16:08:39 -0800 (PST) (envelope-from tamstorf@fa.disney.com) Received: from bacon.fan.fa.disney.com (bacon.fan.fa.disney.com [153.7.118.228]) by plio.fan.fa.disney.com (8.9.2/8.9.2) with ESMTP id QAA27883; Fri, 1 Dec 2000 16:08:38 -0800 (PST) (envelope-from tamstorf@fa.disney.com) Date: Fri, 1 Dec 2000 16:08:38 -0800 From: Rasmus Tamstorf X-Sender: tamstorf@bacon.fan.fa.disney.com To: "J . A . Magallon" cc: OpenInventor Devel List Subject: Re: [PATCH] Linux Makefiles In-Reply-To: <20001202004914.E3982@werewolf.able.es> Message-Id: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing On Sat, 2 Dec 2000, J . A . Magallon wrote: > Hi everyone. > > I have been working to compile OpenInventor on a Linux Mandrake 7.2 - Cooker > system. I have made some changes to Makefiles in order to make them more > readable or suit better to the linux file system tree specs. Patch is attached. > I think it can be included in the CVS tree. > > Changes affect make/ivcommonrules, make/commondefs and make/ivcommondefs. > > Changes: > > - Use 'make -Cdir' instead of '(cd $$i && $(MAKE))'. Less shell spawning. This > can only be used with GNU make (IRIX make has not the -C flag), but as the > README.FIRST file says it has to be built wi gmake... > - If use make -Cdir, don't test if dir exists. If it is not there, make will > complain and give the error msg that now is done manually. Also informs > about the directory change ('entering....' and 'leaving...'). > - Changed test-dir-and-make-it from > if [ ! -d dir ]; then mkdir -p dir; fi > to > mkdir -p dir > This also works on IRIX. If the dir exists, mkdir -p silently ignores, so > no test needed. Also mkdir -p can handle multiple arguments. > - bzip2'ing of man pages. bzip2 is not widely available on IRIX machines, so I don't think this should be the default action. Maybe an option ... > - location of files. In linux there should not be any /usr/demos. All goes in > /usr/share. Specific changes are: Isn't this somewhat of a religious issue ? Some like it in /usr/man others like it in /usr/share/man ... It almost seems better if $(IVROOT) would specify everything up until /man. That way you could set it to '/usr' or '/usr/share' according to taste, but maybe that will break other things. Rasmus ----------------------------------------------------------------------------- Rasmus.Tamstorf@disney.com "A problem worthy of attack, Walt Disney Feature Animation proves its worth by hitting back" Kumbel ----------------------------------------------------------------------------- From owner-info-inventor-dev@oss.sgi.com Fri Dec 1 16:10:23 2000 Received: by oss.sgi.com id ; Fri, 1 Dec 2000 16:10:03 -0800 Received: from jalon.able.es ([212.97.163.2]:34270 "EHLO jalon.able.es") by oss.sgi.com with ESMTP id ; Fri, 1 Dec 2000 16:09:58 -0800 Received: from correo.able.es ([212.97.169.121]) by jalon.able.es (Netscape Messaging Server 4.15) with SMTP id G4WZAS00.P5U for ; Sat, 2 Dec 2000 01:13:40 +0100 Date: Sat, 2 Dec 2000 01:09:46 +0100 From: "J . A . Magallon" To: OpenInventor Devel List Subject: [PATCH] GL12 Compatibility Message-ID: <20001202010946.I3982@werewolf.able.es> Reply-To: jamagallon@able.es Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="+fexcqMh/evT6CrY" Content-Transfer-Encoding: 8bit X-Mailer: Balsa 1.0.0 Content-Length: 8335 Lines: 280 Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing --+fexcqMh/evT6CrY Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Hi everyone, I have been building OIV in a Mandrake 7.2 with XFree4.0.1. This xfree includes version 3.4 of Mesa, that is supposed to be GL1.2 compatible. That gives some drawbacks whe building Open Inventor. Basicly, inventor is coded against old OpenGL 1.0, where (for example) TexObjs were an extension, and recent GLs are 1.1 or 1.2, where TexObjs ar mandatory and functions do not have the EXT tail. Some implementations can support both interfaces (I think even mesa can do through preprocessor define flags), but I thin it is cleaner to write the code like: #ifndef GL_VERSION_1_1 glBindTextureEXT(GL_TEXTURE_2D, 0); #else glBindTexture(GL_TEXTURE_2D, 0); #endif I also think that for this kind of features (EXTs that have been adopted) one could skip the tests for extension presence, because the interface is supposed to be usable even if it does nothing (ie, it is just a software cache...), but I am not fully confident and so have made the minimal changes. If someone can tell me sure (hey, SGI people...) that this can be made, I can prepare the patch. Changes are: - A couple of casts to make g++2.96 shut up. - Poylgon offset interface. - Texture objects - Tesselator interface if GLU version is 1.2. Files affected are: apps/demos/textomatic/TextWrapper.c++ apps/nodes/Decal/Decal.c++ lib/database/src/so/elements/SoGLTextureImageElement.c++ lib/database/src/so/elements/SoGLCacheContextElement.c++ lib/database/src/so/nodes/SoShape.c++ lib/database/src/so/nodes/SoAsciiText.c++ Please, apply. I think it does not break anything and updates build compatibility. If you apply this on CVS, please let me or the list know, so I trash my patches and download corrected CVS. Thank you. -- Juan Antonio Magallon Lacarta #> cd /pub mailto:jamagallon@able.es #> more beer Linux 2.2.18-pre24-vm #2 SMP Wed Nov 29 02:56:21 CET 2000 i686 unknown --+fexcqMh/evT6CrY Content-Type: application/octet-stream; charset=us-ascii Content-Disposition: attachment; filename=patch-gl12 --- inventor/apps/demos/textomatic/TextWrapper.c++.org Fri Nov 24 00:55:53 2000 +++ inventor/apps/demos/textomatic/TextWrapper.c++ Fri Nov 24 00:58:08 2000 @@ -77,7 +77,7 @@ char *output = outbuf; // if ( global_iconvCodeL2 == (iconv_t)-1 || - iconv(global_iconvCodeL2,(const char **)&input,&inbytesleft,&output,&outbytesleft) == (size_t)-1 ) + iconv(global_iconvCodeL2,&input,&inbytesleft,&output,&outbytesleft) == (size_t)-1 ) { fprintf( stderr, "textomatic: iconv error.\n" ); (*(UCS2 *)outbuf)=0; @@ -229,7 +229,7 @@ char *output = outbuf; // if ( global_iconvCode28 == (iconv_t)-1 || - iconv( global_iconvCode28,(const char **)&input,&inbytesleft,&output,&outbytesleft ) == (size_t)-1 ) + iconv( global_iconvCode28,&input,&inbytesleft,&output,&outbytesleft ) == (size_t)-1 ) { fprintf( stderr, "textomatic: iconv error.\n" ); (*(UCS2 *)outbuf)=0; --- inventor/apps/nodes/Decal/Decal.c++.org Fri Nov 24 00:43:18 2000 +++ inventor/apps/nodes/Decal/Decal.c++ Fri Nov 24 00:50:57 2000 @@ -152,18 +152,34 @@ } if (SoGLCacheContextElement::extSupported(state, offsetExtInt)) { +#ifdef GL_VERSION_1_1 + glEnable(GL_POLYGON_OFFSET_FILL); + glEnable(GL_POLYGON_OFFSET_LINE); + glEnable(GL_POLYGON_OFFSET_POINT); +#else glEnable(GL_POLYGON_OFFSET_EXT); +#endif float offset = 0; for (int i=0; i < childList->getLength(); i++) { +#ifdef GL_VERSION_1_1 + glPolygonOffset(offset, 0); +#else glPolygonOffsetEXT(offset, 0); +#endif childList->traverse(action, i); offset -= 0.1; } +#ifdef GL_VERSION_1_1 + glDisable(GL_POLYGON_OFFSET_FILL); + glDisable(GL_POLYGON_OFFSET_LINE); + glDisable(GL_POLYGON_OFFSET_POINT); +#else glDisable(GL_POLYGON_OFFSET_EXT); +#endif } else { --- inventor/lib/database/src/so/elements/SoGLTextureImageElement.c++.org Sun Nov 19 01:37:59 2000 +++ inventor/lib/database/src/so/elements/SoGLTextureImageElement.c++ Sun Nov 19 01:38:46 2000 @@ -407,7 +407,11 @@ // unbind the current texture object so we don't overwrite it's state. #ifdef GL_EXT_texture_object if (!buildList) +#ifndef GL_VERSION_1_1 glBindTextureEXT(GL_TEXTURE_2D, 0); +#else + glBindTexture(GL_TEXTURE_2D, 0); +#endif #endif // These need to go inside the display list or texture object --- inventor/lib/database/src/so/elements/SoGLCacheContextElement.c++.org Sun Nov 19 01:24:54 2000 +++ inventor/lib/database/src/so/elements/SoGLCacheContextElement.c++ Sun Nov 19 01:37:16 2000 @@ -437,7 +437,11 @@ if (type == TEXTURE_OBJECT) { #ifdef GL_EXT_texture_object +#if !defined(GL_VERSION_1_1) glGenTexturesEXT(1, &startIndex); +#else + glGenTextures(1, &startIndex); +#endif #ifdef DEBUG if (num != 1) SoDebugError::post("SoGLDisplayList", "Sorry, can only " @@ -502,7 +506,11 @@ { if (type == TEXTURE_OBJECT) { #ifdef GL_EXT_texture_object +#if !defined(GL_VERSION_1_1) glBindTextureEXT(GL_TEXTURE_2D, startIndex+index); +#else + glBindTexture(GL_TEXTURE_2D, startIndex+index); +#endif #endif } else { glNewList(startIndex+index, GL_COMPILE_AND_EXECUTE); @@ -540,7 +548,11 @@ { if (type == TEXTURE_OBJECT) { #ifdef GL_EXT_texture_object +#if !defined(GL_VERSION_1_1) glBindTextureEXT(GL_TEXTURE_2D, startIndex+index); +#else + glBindTexture(GL_TEXTURE_2D, startIndex+index); +#endif #endif } else { glCallList(startIndex+index); @@ -580,7 +592,11 @@ { if (type == TEXTURE_OBJECT) { #ifdef GL_EXT_texture_object +#if !defined(GL_VERSION_1_1) glDeleteTexturesEXT(1, &startIndex); +#else + glDeleteTextures(1, &startIndex); +#endif #endif } else { glDeleteLists(startIndex, num); --- inventor/lib/database/src/so/nodes/SoShape.c++.org Sun Nov 19 23:45:15 2000 +++ inventor/lib/database/src/so/nodes/SoShape.c++ Sun Nov 19 23:59:11 2000 @@ -886,7 +886,12 @@ gluTessCallback(tobj, (GLenum)GLU_ERROR, (void (*)())SoShape::errorCB); } +#ifndef GLU_VERSION_1_2 gluBeginPolygon(tobj); +#else + gluTessBeginPolygon(tobj,0); + gluTessBeginContour(tobj); +#endif for (i = 0; i < polyVertNum; i++) { const SbVec3f &t = polyVerts[i].getPoint(); @@ -895,7 +900,12 @@ dv[0] = t[0]; dv[1] = t[1]; dv[2] = t[2]; gluTessVertex(tobj, dv, (void *)&polyVerts[i]); } +#ifndef GLU_VERSION_1_2 gluEndPolygon(tobj); +#else + gluTessEndContour(tobj); + gluTessEndPolygon(tobj); +#endif polyVertNum = 0; break; --- inventor/lib/database/src/so/nodes/SoAsciiText.c++.org Sun Nov 19 01:54:16 2000 +++ inventor/lib/database/src/so/nodes/SoAsciiText.c++ Sun Nov 19 01:59:40 2000 @@ -1389,7 +1389,12 @@ GLdouble v[3]; tesselationError = FALSE; +#ifndef GLU_VERSION_1_2 gluBeginPolygon(tobj); +#else + gluTessBeginPolygon(tobj,0); + gluTessBeginContour(tobj); +#endif // Get outline for character MyFontOutline *outline = getOutline(c); @@ -1398,7 +1403,12 @@ // It would be nice if the font manager told us the type of // each outline... +#ifndef GLU_VERSION_1_2 gluNextContour(tobj, (GLenum)GLU_UNKNOWN); +#else + gluTessEndContour(tobj); + gluTessBeginContour(tobj); +#endif for (int j = 0; j < outline->getNumVerts(i); j++) { SbVec2f &t = outline->getVertex(i,j); @@ -1412,7 +1422,12 @@ gluTessVertex(tobj, v, &t); } } +#ifndef GLU_VERSION_1_2 gluEndPolygon(tobj); +#else + gluTessEndContour(tobj); + gluTessEndPolygon(tobj); +#endif // If there was an error tesselating the character, just generate // a bounding box for the character: @@ -1425,14 +1440,24 @@ boxVerts[1].setValue(boxVerts[2][0], boxVerts[0][1]); boxVerts[3].setValue(boxVerts[0][0], boxVerts[2][1]); +#ifndef GLU_VERSION_1_2 gluBeginPolygon(tobj); +#else + gluTessBeginPolygon(tobj,0); + gluTessBeginContour(tobj); +#endif for (i = 0; i < 4; i++) { v[0] = boxVerts[i][0]; v[1] = boxVerts[i][1]; v[2] = 0.0; gluTessVertex(tobj, v, &boxVerts[i]); } +#ifndef GLU_VERSION_1_2 gluEndPolygon(tobj); +#else + gluTessEndContour(tobj); + gluTessEndPolygon(tobj); +#endif } } } --+fexcqMh/evT6CrY-- From owner-info-inventor-dev@oss.sgi.com Fri Dec 1 16:16:03 2000 Received: by oss.sgi.com id ; Fri, 1 Dec 2000 16:15:53 -0800 Received: from pneumatic-tube.sgi.com ([204.94.214.22]:25358 "EHLO pneumatic-tube.sgi.com") by oss.sgi.com with ESMTP id ; Fri, 1 Dec 2000 16:15:42 -0800 Received: from kinabalu.csd.sgi.com (kinabalu.csd.sgi.com [150.166.1.88]) by pneumatic-tube.sgi.com (980327.SGI.8.8.8-aspam/980310.SGI-aspam) via ESMTP id QAA05647 for ; Fri, 1 Dec 2000 16:23:46 -0800 (PST) mail_from (jlim@kinabalu.csd.sgi.com) Received: by kinabalu.csd.sgi.com (980427.SGI.8.8.8/911001.SGI) id QAA35175; Fri, 1 Dec 2000 16:13:34 -0800 (PST) From: jlim@kinabalu.csd.sgi.com (Jonathan Lim) Message-Id: <200012020013.QAA35175@kinabalu.csd.sgi.com> Subject: Re: [PATCH] Linux Makefiles To: tamstorf@fa.disney.com (Rasmus Tamstorf) Date: Fri, 1 Dec 2000 16:13:34 -0800 (PST) Cc: jamagallon@able.es (J . A . Magallon), info-inventor-dev@oss.sgi.com (OpenInventor Devel List) In-Reply-To: from "Rasmus Tamstorf" at Dec 01, 2000 04:08:38 PM X-Mailer: ELM [version 2.5 PL2] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing On Fri Dec 1 16:08:38 2000, tamstorf@fa.disney.com wrote: > > > Changes: > > > > - Use 'make -Cdir' instead of '(cd $$i && $(MAKE))'. Less shell spawning. This > > can only be used with GNU make (IRIX make has not the -C flag), but as the > > README.FIRST file says it has to be built wi gmake... This change is not a problem. > > - bzip2'ing of man pages. > > bzip2 is not widely available on IRIX machines, so I don't think this > should be the default action. Maybe an option ... Agreed. > > - location of files. In linux there should not be any /usr/demos. All goes in > > /usr/share. Specific changes are: > > Isn't this somewhat of a religious issue ? Some like it in /usr/man others > like it in /usr/share/man ... It almost seems better if $(IVROOT) would > specify everything up until /man. That way you could set it to '/usr' or > '/usr/share' according to taste, but maybe that will break other things. My thoughts exactly. Besides, some of the viewers and demos have hardcoded the paths to the help files. > Rasmus Jonathan Lim _ Silicon Graphics _ Mountain View GPS Graphics Computer Systems CA, USA From owner-info-inventor-dev@oss.sgi.com Fri Dec 1 16:16:53 2000 Received: by oss.sgi.com id ; Fri, 1 Dec 2000 16:16:33 -0800 Received: from jalon.able.es ([212.97.163.2]:41694 "EHLO jalon.able.es") by oss.sgi.com with ESMTP id ; Fri, 1 Dec 2000 16:16:21 -0800 Received: from correo.able.es ([212.97.169.121]) by jalon.able.es (Netscape Messaging Server 4.15) with SMTP id G4WZLH00.04Q; Sat, 2 Dec 2000 01:20:05 +0100 Date: Sat, 2 Dec 2000 01:16:12 +0100 From: "J . A . Magallon" To: Rasmus Tamstorf Cc: OpenInventor Devel List Subject: Re: [PATCH] Linux Makefiles Message-ID: <20001202011612.K3982@werewolf.able.es> Reply-To: jamagallon@able.es References: <20001202004914.E3982@werewolf.able.es> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In-Reply-To: ; from tamstorf@fa.disney.com on Sat, Dec 02, 2000 at 01:08:38 +0100 X-Mailer: Balsa 1.0.0 Content-Length: 1241 Lines: 25 Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing On Sat, 02 Dec 2000 01:08:38 Rasmus Tamstorf wrote: > On Sat, 2 Dec 2000, J . A . Magallon wrote: > > Isn't this somewhat of a religious issue ? Some like it in /usr/man others > like it in /usr/share/man ... It almost seems better if $(IVROOT) would > specify everything up until /man. That way you could set it to '/usr' or > '/usr/share' according to taste, but maybe that will break other things. I think it is related to some standard (something like the Linux File System Standard), that everybody is going to. I think RedHat uses /usr/share for man, and Mandrake surely does that, it changed between 7.1 and 7.2 (the 'Big Move'). There are other issues like /opt instead of /usr/local and so on. But nowadays the most annoyig are the moves from /usr/man to /usr/share/man, and from /usr/doc to /usr/share/doc. You still found rpms that put their docs in the old place... I think all that is to mount /usr/share via NFS and have there all the 'shared' non binary data, instead of mounting man, doc, etc... -- Juan Antonio Magallon Lacarta #> cd /pub mailto:jamagallon@able.es #> more beer Linux 2.2.18-pre24-vm #2 SMP Wed Nov 29 02:56:21 CET 2000 i686 unknown From owner-info-inventor-dev@oss.sgi.com Fri Dec 1 16:21:53 2000 Received: by oss.sgi.com id ; Fri, 1 Dec 2000 16:21:33 -0800 Received: from jalon.able.es ([212.97.163.2]:47326 "EHLO jalon.able.es") by oss.sgi.com with ESMTP id ; Fri, 1 Dec 2000 16:21:18 -0800 Received: from correo.able.es ([212.97.169.121]) by jalon.able.es (Netscape Messaging Server 4.15) with SMTP id G4WZTB00.T6K; Sat, 2 Dec 2000 01:24:47 +0100 Date: Sat, 2 Dec 2000 01:20:53 +0100 From: "J . A . Magallon" To: Jonathan Lim Cc: Rasmus Tamstorf , "J . A . Magallon" , OpenInventor Devel List Subject: Re: [PATCH] Linux Makefiles Message-ID: <20001202012053.M3982@werewolf.able.es> Reply-To: jamagallon@able.es References: <200012020013.QAA35175@kinabalu.csd.sgi.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In-Reply-To: <200012020013.QAA35175@kinabalu.csd.sgi.com>; from jlim@kinabalu.csd.sgi.com on Sat, Dec 02, 2000 at 01:13:34 +0100 X-Mailer: Balsa 1.0.0 Content-Length: 1315 Lines: 35 Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing On Sat, 02 Dec 2000 01:13:34 Jonathan Lim wrote: > > bzip2 is not widely available on IRIX machines, so I don't think this > > should be the default action. Maybe an option ... > > Agreed. Do I cut the patch or do it yourself ? (I understand it is too Mandrake specific) > > > > - location of files. In linux there should not be any /usr/demos. All goes > in > > > /usr/share. Specific changes are: > > > > Isn't this somewhat of a religious issue ? Some like it in /usr/man others > > like it in /usr/share/man ... It almost seems better if $(IVROOT) would > > specify everything up until /man. That way you could set it to '/usr' or > > '/usr/share' according to taste, but maybe that will break other things. > > My thoughts exactly. Besides, some of the viewers and demos have hardcoded the > paths to the help files. > But that should be changed. If you port a package, you have to make the package behave like a Linux package, not perturb the standard linux file system to make it look like an SGI. If not, you can end like Microsoft puttin DLLs inside an SGI to run XSI. -- Juan Antonio Magallon Lacarta #> cd /pub mailto:jamagallon@able.es #> more beer Linux 2.2.18-pre24-vm #2 SMP Wed Nov 29 02:56:21 CET 2000 i686 unknown From owner-info-inventor-dev@oss.sgi.com Fri Dec 1 16:28:53 2000 Received: by oss.sgi.com id ; Fri, 1 Dec 2000 16:28:33 -0800 Received: from pneumatic-tube.sgi.com ([204.94.214.22]:3855 "EHLO pneumatic-tube.sgi.com") by oss.sgi.com with ESMTP id ; Fri, 1 Dec 2000 16:28:19 -0800 Received: from kinabalu.csd.sgi.com (kinabalu.csd.sgi.com [150.166.1.88]) by pneumatic-tube.sgi.com (980327.SGI.8.8.8-aspam/980310.SGI-aspam) via ESMTP id QAA05632 for ; Fri, 1 Dec 2000 16:36:23 -0800 (PST) mail_from (jlim@kinabalu.csd.sgi.com) Received: by kinabalu.csd.sgi.com (980427.SGI.8.8.8/911001.SGI) id QAA37932; Fri, 1 Dec 2000 16:26:09 -0800 (PST) From: jlim@kinabalu.csd.sgi.com (Jonathan Lim) Message-Id: <200012020026.QAA37932@kinabalu.csd.sgi.com> Subject: Re: [PATCH] Linux Makefiles To: jamagallon@able.es Date: Fri, 1 Dec 2000 16:26:08 -0800 (PST) Cc: tamstorf@fa.disney.com (Rasmus Tamstorf), info-inventor-dev@oss.sgi.com (OpenInventor Devel List) In-Reply-To: <20001202012053.M3982@werewolf.able.es> from "J . A . Magallon" at Dec 02, 2000 01:20:53 AM X-Mailer: ELM [version 2.5 PL2] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing On Fri Dec 1 16:20:53 2000, jamagallon@able.es wrote: > > On Sat, 02 Dec 2000 01:13:34 Jonathan Lim wrote: > > > bzip2 is not widely available on IRIX machines, so I don't think this > > > should be the default action. Maybe an option ... > > > > Agreed. > > Do I cut the patch or do it yourself ? (I understand it is too Mandrake > specific) Let me think about it a bit first. > But that should be changed. If you port a package, you have to make the > package behave like a Linux package, not perturb the standard linux file > system to make it look like an SGI. > > If not, you can end like Microsoft puttin DLLs inside an SGI to run XSI. Speaking of Microsoft, someone recently submitted a whole lot of changes to get things to work on Windows. I haven't even begun to look into it yet, have you? Jonathan Lim _ Silicon Graphics _ Mountain View GPS Graphics Computer Systems CA, USA From owner-info-inventor-dev@oss.sgi.com Fri Dec 1 16:36:23 2000 Received: by oss.sgi.com id ; Fri, 1 Dec 2000 16:36:04 -0800 Received: from jalon.able.es ([212.97.163.2]:62942 "EHLO jalon.able.es") by oss.sgi.com with ESMTP id ; Fri, 1 Dec 2000 16:35:54 -0800 Received: from correo.able.es ([212.97.169.121]) by jalon.able.es (Netscape Messaging Server 4.15) with SMTP id G4X0I400.A4S; Sat, 2 Dec 2000 01:39:40 +0100 Date: Sat, 2 Dec 2000 01:35:46 +0100 From: "J . A . Magallon" To: Jonathan Lim Cc: OpenInventor Devel List Subject: Re: [PATCH] Linux Makefiles Message-ID: <20001202013546.A6767@werewolf.able.es> Reply-To: jamagallon@able.es References: <20001202012053.M3982@werewolf.able.es> <200012020026.QAA37932@kinabalu.csd.sgi.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In-Reply-To: <200012020026.QAA37932@kinabalu.csd.sgi.com>; from jlim@kinabalu.csd.sgi.com on Sat, Dec 02, 2000 at 01:26:08 +0100 X-Mailer: Balsa 1.0.0 Content-Length: 766 Lines: 20 Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing On Sat, 02 Dec 2000 01:26:08 Jonathan Lim wrote: > On Fri Dec 1 16:20:53 2000, jamagallon@able.es wrote: > > > Speaking of Microsoft, someone recently submitted a whole lot of changes to > get > things to work on Windows. I haven't even begun to look into it yet, have you? > No, I have not seen that. Was this list or the inventor-info ? Anyways, for example, I just saw the first answer for the subject on GLwMDrawA, no the original mail, so perhaps I am not receiving all the mail in the list. Gotta subscribe the other list also... -- Juan Antonio Magallon Lacarta #> cd /pub mailto:jamagallon@able.es #> more beer Linux 2.2.18-pre24-vm #2 SMP Wed Nov 29 02:56:21 CET 2000 i686 unknown From owner-info-inventor-dev@oss.sgi.com Fri Dec 1 16:40:23 2000 Received: by oss.sgi.com id ; Fri, 1 Dec 2000 16:40:03 -0800 Received: from mail.disney.com ([204.128.192.15]:33212 "EHLO mail.disney.com") by oss.sgi.com with ESMTP id ; Fri, 1 Dec 2000 16:39:52 -0800 Received: from pain10.corp.disney.com (root@pain10.corp.disney.com [153.7.110.100]) by mail.disney.com (Switch-2.0.1/Switch-2.0.1) with SMTP id eB20dkp17186 for ; Fri, 1 Dec 2000 16:39:46 -0800 (PST) Received: from louie.fa.disney.com by pain.corp.disney.com with ESMTP; Fri, 1 Dec 2000 16:40:31 -0800 Received: from plio.fan.fa.disney.com (plio.fan.fa.disney.com [153.7.118.2]) by louie.fa.disney.com (8.9.2/8.9.2) with ESMTP id QAA18786; Fri, 1 Dec 2000 16:39:50 -0800 (PST) (envelope-from tamstorf@fa.disney.com) Received: from bacon.fan.fa.disney.com (bacon.fan.fa.disney.com [153.7.118.228]) by plio.fan.fa.disney.com (8.9.2/8.9.2) with ESMTP id QAA29690; Fri, 1 Dec 2000 16:39:50 -0800 (PST) (envelope-from tamstorf@fa.disney.com) Date: Fri, 1 Dec 2000 16:39:49 -0800 From: Rasmus Tamstorf X-Sender: tamstorf@bacon.fan.fa.disney.com To: "J . A . Magallon" cc: Jonathan Lim , OpenInventor Devel List Subject: Re: [PATCH] Linux Makefiles In-Reply-To: <20001202012053.M3982@werewolf.able.es> Message-Id: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing On Sat, 2 Dec 2000, J . A . Magallon wrote: > > > > - location of files. In linux there should not be any /usr/demos. All goes > > in > > > > /usr/share. Specific changes are: > > > > > > Isn't this somewhat of a religious issue ? Some like it in /usr/man others > > > like it in /usr/share/man ... It almost seems better if $(IVROOT) would > > > specify everything up until /man. That way you could set it to '/usr' or > > > '/usr/share' according to taste, but maybe that will break other things. > > > > My thoughts exactly. Besides, some of the viewers and demos have hardcoded the > > paths to the help files. > > > > But that should be changed. If you port a package, you have to make the > package behave like a Linux package, not perturb the standard linux file > system to make it look like an SGI. > > If not, you can end like Microsoft puttin DLLs inside an SGI to run XSI. Agreed, but as the name indicates, it's *Open* inventor. Not Linux Inventor. Whatever is done should be flexible enough to support both Linux and IRIX as well as possibly other OS'es. Hardcoding for Linux is no better than your Microsoft example. Rasmus ----------------------------------------------------------------------------- Rasmus.Tamstorf@disney.com "A problem worthy of attack, Walt Disney Feature Animation proves its worth by hitting back" Kumbel ----------------------------------------------------------------------------- From owner-info-inventor-dev@oss.sgi.com Fri Dec 1 16:47:03 2000 Received: by oss.sgi.com id ; Fri, 1 Dec 2000 16:46:54 -0800 Received: from jalon.able.es ([212.97.163.2]:9951 "EHLO jalon.able.es") by oss.sgi.com with ESMTP id ; Fri, 1 Dec 2000 16:46:41 -0800 Received: from correo.able.es ([212.97.169.121]) by jalon.able.es (Netscape Messaging Server 4.15) with SMTP id G4X10300.853; Sat, 2 Dec 2000 01:50:27 +0100 Date: Sat, 2 Dec 2000 01:46:33 +0100 From: "J . A . Magallon" To: Rasmus Tamstorf Cc: OpenInventor Devel List Subject: Re: [PATCH] Linux Makefiles Message-ID: <20001202014633.G6767@werewolf.able.es> Reply-To: jamagallon@able.es References: <20001202012053.M3982@werewolf.able.es> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In-Reply-To: ; from tamstorf@fa.disney.com on Sat, Dec 02, 2000 at 01:39:49 +0100 X-Mailer: Balsa 1.0.0 Content-Length: 715 Lines: 18 Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing On Sat, 02 Dec 2000 01:39:49 Rasmus Tamstorf wrote: > > Agreed, but as the name indicates, it's *Open* inventor. Not Linux > Inventor. Whatever is done should be flexible enough to support both > Linux and IRIX as well as possibly other OS'es. Hardcoding for Linux is > no better than your Microsoft example. > I agree with that. I only see these things as temporary while somebody i read in the list finishes his try with autoconf/automake. That handles all this issues the right way. -- Juan Antonio Magallon Lacarta #> cd /pub mailto:jamagallon@able.es #> more beer Linux 2.2.18-pre24-vm #2 SMP Wed Nov 29 02:56:21 CET 2000 i686 unknown From owner-info-inventor-dev@oss.sgi.com Tue Dec 12 13:17:46 2000 Received: by oss.sgi.com id ; Tue, 12 Dec 2000 13:17:26 -0800 Received: from client82-225.hispeed.ch ([62.2.82.225]:28933 "EHLO clochette.net2000.ch") by oss.sgi.com with ESMTP id ; Tue, 12 Dec 2000 13:16:53 -0800 Received: (from erik@localhost) by clochette.net2000.ch (8.10.2/8.10.2/SuSE Linux 8.10.0-0.3) id eBCLHrM01559 for info-inventor-dev@oss.sgi.com; Tue, 12 Dec 2000 22:17:53 +0100 Date: Tue, 12 Dec 2000 22:17:53 +0100 From: Erik Simon To: info-inventor-dev@oss.sgi.com Subject: Text does not work in Linux (ATI/glx) Message-ID: <20001212221753.A1526@localhost> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing Hi, I'm using OpenInventor 2.1 on Linux : OpenInventor rpms: sgi-OpenInventor-clients-2.1.5-6 sgi-OpenInventor-data-2.1.5-6 sgi-OpenInventor-devel-2.1.5-6 Linux: SuSE 7.0 (kernel 2.2.15) XFree 3.3.6 (xf86_3x-3.3.6-47.rpm) GLX (glx-000623-1.rpm) Mesa 3.2 (mesa-3.2-63.rpm) Hardware: PII 400 ATI 3D RagePro AGP (used as PCI by XF86 with driver Mach64) Note that I had to disable pixmap/font caches to be able to use glx. The text (SoText3) does not appear in my programs. Here is what I get: @@Created GLX Context.. invalid data type to GLX_data_size: 0x1407!! invalid data type to GLX_data_size: 0x1407!! invalid data type to GLX_data_size: 0x1407!! invalid data type to GLX_data_size: 0x1407!! I got the same error when I tried to redirect the display from a SGI running the same code to my Linux display. I guess there might be a bug in GLX, since the text works fine with Mesasoft, but this doesn't use the hardware acceleration of my card, which is sad. Any idea? Thanks. Erik Simon erik.simon@unine.ch From owner-info-inventor-dev@oss.sgi.com Tue Dec 12 18:03:59 2000 Received: by oss.sgi.com id ; Tue, 12 Dec 2000 18:03:39 -0800 Received: from pneumatic-tube.sgi.com ([204.94.214.22]:11887 "EHLO pneumatic-tube.sgi.com") by oss.sgi.com with ESMTP id ; Tue, 12 Dec 2000 18:03:27 -0800 Received: from kinabalu.csd.sgi.com (kinabalu.csd.sgi.com [150.166.1.88]) by pneumatic-tube.sgi.com (980327.SGI.8.8.8-aspam/980310.SGI-aspam) via ESMTP id SAA02924 for ; Tue, 12 Dec 2000 18:11:41 -0800 (PST) mail_from (jlim@kinabalu.csd.sgi.com) Received: by kinabalu.csd.sgi.com (980427.SGI.8.8.8/911001.SGI) id SAA89819; Tue, 12 Dec 2000 18:01:10 -0800 (PST) From: jlim@kinabalu.csd.sgi.com (Jonathan Lim) Message-Id: <200012130201.SAA89819@kinabalu.csd.sgi.com> Subject: Re: [PATCH] GL12 Compatibility To: jamagallon@able.es Date: Tue, 12 Dec 2000 18:01:10 -0800 (PST) Cc: info-inventor-dev@oss.sgi.com (OpenInventor Devel List) In-Reply-To: <20001202010946.I3982@werewolf.able.es> from "J . A . Magallon" at Dec 02, 2000 01:09:46 AM X-Mailer: ELM [version 2.5 PL2] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing Hi all, I'm about the check the following changes in. Has anyone else encountered similar problems and wish to add to the discussion? Jonathan Lim _ Silicon Graphics _ Mountain View GPS Graphics Computer Systems CA, USA On Fri Dec 1 16:09:46 2000, jamagallon@able.es wrote: > > Hi everyone, > > I have been building OIV in a Mandrake 7.2 with XFree4.0.1. This xfree includes > version 3.4 of Mesa, that is supposed to be GL1.2 compatible. That gives some > drawbacks whe building Open Inventor. Basicly, inventor is coded against > old OpenGL 1.0, where (for example) TexObjs were an extension, and recent > GLs are 1.1 or 1.2, where TexObjs ar mandatory and functions do not have the > EXT tail. > > Some implementations can support both interfaces (I think even mesa can do > through preprocessor define flags), but I thin it is cleaner to write the > code like: > > #ifndef GL_VERSION_1_1 > glBindTextureEXT(GL_TEXTURE_2D, 0); > #else > glBindTexture(GL_TEXTURE_2D, 0); > #endif > > I also think that for this kind of features (EXTs that have been adopted) > one could skip the tests for extension presence, because the interface > is supposed to be usable even if it does nothing (ie, it is just a > software cache...), but I am not fully confident and so have made the > minimal changes. If someone can tell me sure (hey, SGI people...) that > this can be made, I can prepare the patch. > > Changes are: > > - A couple of casts to make g++2.96 shut up. > - Poylgon offset interface. > - Texture objects > - Tesselator interface if GLU version is 1.2. > > Files affected are: > apps/demos/textomatic/TextWrapper.c++ > apps/nodes/Decal/Decal.c++ > lib/database/src/so/elements/SoGLTextureImageElement.c++ > lib/database/src/so/elements/SoGLCacheContextElement.c++ > lib/database/src/so/nodes/SoShape.c++ > lib/database/src/so/nodes/SoAsciiText.c++ > > Please, apply. I think it does not break anything and updates build > compatibility. > > If you apply this on CVS, please let me or the list know, so I trash > my patches and download corrected CVS. > > Thank you. > > -- > Juan Antonio Magallon Lacarta #> cd /pub > mailto:jamagallon@able.es #> more beer From owner-info-inventor-dev@oss.sgi.com Wed Dec 13 21:00:27 2000 Received: by oss.sgi.com id ; Wed, 13 Dec 2000 21:00:09 -0800 Received: from [195.158.3.33] ([195.158.3.33]:31750 "EHLO nuron.ccc.uz") by oss.sgi.com with ESMTP id ; Wed, 13 Dec 2000 20:59:59 -0800 Received: (from root@localhost) by nuron.ccc.uz (8.10.2/8.10.2) id eBE536N24470; Thu, 14 Dec 2000 10:03:06 +0500 Received: from oss.sgi.com (IDENT:root@oss.sgi.com [216.32.174.190]) by nuron.ccc.uz (8.10.2/8.10.2) with ESMTP id eBD27mH06380 for ; Wed, 13 Dec 2000 07:07:49 +0500 Received: by oss.sgi.com id ; Tue, 12 Dec 2000 18:03:39 -0800 Received: from pneumatic-tube.sgi.com ([204.94.214.22]:11887 "EHLO pneumatic-tube.sgi.com") by oss.sgi.com with ESMTP id ; Tue, 12 Dec 2000 18:03:27 -0800 Received: from kinabalu.csd.sgi.com (kinabalu.csd.sgi.com [150.166.1.88]) by pneumatic-tube.sgi.com (980327.SGI.8.8.8-aspam/980310.SGI-aspam) via ESMTP id SAA02924 for ; Tue, 12 Dec 2000 18:11:41 -0800 (PST) mail_from (jlim@kinabalu.csd.sgi.com) Received: by kinabalu.csd.sgi.com (980427.SGI.8.8.8/911001.SGI) id SAA89819; Tue, 12 Dec 2000 18:01:10 -0800 (PST) From: jlim@kinabalu.csd.sgi.com (Jonathan Lim) Message-Id: <200012130201.SAA89819@kinabalu.csd.sgi.com> Subject: Re: [PATCH] GL12 Compatibility To: jamagallon@able.es Date: Tue, 12 Dec 2000 18:01:10 -0800 (PST) Cc: info-inventor-dev@oss.sgi.com (OpenInventor Devel List) In-Reply-To: <20001202010946.I3982@werewolf.able.es> from "J . A . Magallon" at Dec 02, 2000 01:09:46 AM X-Mailer: ELM [version 2.5 PL2] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing Hi all, I'm about the check the following changes in. Has anyone else encountered similar problems and wish to add to the discussion? Jonathan Lim _ Silicon Graphics _ Mountain View GPS Graphics Computer Systems CA, USA On Fri Dec 1 16:09:46 2000, jamagallon@able.es wrote: > > Hi everyone, > > I have been building OIV in a Mandrake 7.2 with XFree4.0.1. This xfree includes > version 3.4 of Mesa, that is supposed to be GL1.2 compatible. That gives some > drawbacks whe building Open Inventor. Basicly, inventor is coded against > old OpenGL 1.0, where (for example) TexObjs were an extension, and recent > GLs are 1.1 or 1.2, where TexObjs ar mandatory and functions do not have the > EXT tail. > > Some implementations can support both interfaces (I think even mesa can do > through preprocessor define flags), but I thin it is cleaner to write the > code like: > > #ifndef GL_VERSION_1_1 > glBindTextureEXT(GL_TEXTURE_2D, 0); > #else > glBindTexture(GL_TEXTURE_2D, 0); > #endif > > I also think that for this kind of features (EXTs that have been adopted) > one could skip the tests for extension presence, because the interface > is supposed to be usable even if it does nothing (ie, it is just a > software cache...), but I am not fully confident and so have made the > minimal changes. If someone can tell me sure (hey, SGI people...) that > this can be made, I can prepare the patch. > > Changes are: > > - A couple of casts to make g++2.96 shut up. > - Poylgon offset interface. > - Texture objects > - Tesselator interface if GLU version is 1.2. > > Files affected are: > apps/demos/textomatic/TextWrapper.c++ > apps/nodes/Decal/Decal.c++ > lib/database/src/so/elements/SoGLTextureImageElement.c++ > lib/database/src/so/elements/SoGLCacheContextElement.c++ > lib/database/src/so/nodes/SoShape.c++ > lib/database/src/so/nodes/SoAsciiText.c++ > > Please, apply. I think it does not break anything and updates build > compatibility. > > If you apply this on CVS, please let me or the list know, so I trash > my patches and download corrected CVS. > > Thank you. > > -- > Juan Antonio Magallon Lacarta #> cd /pub > mailto:jamagallon@able.es #> more beer From owner-info-inventor-dev@oss.sgi.com Wed Dec 13 21:00:28 2000 Received: by oss.sgi.com id ; Wed, 13 Dec 2000 21:00:08 -0800 Received: from [195.158.3.33] ([195.158.3.33]:27398 "EHLO nuron.ccc.uz") by oss.sgi.com with ESMTP id ; Wed, 13 Dec 2000 20:59:49 -0800 Received: (from root@localhost) by nuron.ccc.uz (8.10.2/8.10.2) id eBE530r24453; Thu, 14 Dec 2000 10:03:00 +0500 Received: from oss.sgi.com (IDENT:root@oss.sgi.com [216.32.174.190]) by nuron.ccc.uz (8.10.2/8.10.2) with ESMTP id eBCLLfH04836 for ; Wed, 13 Dec 2000 02:21:42 +0500 Received: by oss.sgi.com id ; Tue, 12 Dec 2000 13:17:26 -0800 Received: from client82-225.hispeed.ch ([62.2.82.225]:28933 "EHLO clochette.net2000.ch") by oss.sgi.com with ESMTP id ; Tue, 12 Dec 2000 13:16:53 -0800 Received: (from erik@localhost) by clochette.net2000.ch (8.10.2/8.10.2/SuSE Linux 8.10.0-0.3) id eBCLHrM01559 for info-inventor-dev@oss.sgi.com; Tue, 12 Dec 2000 22:17:53 +0100 Date: Tue, 12 Dec 2000 22:17:53 +0100 From: Erik Simon To: info-inventor-dev@oss.sgi.com Subject: Text does not work in Linux (ATI/glx) Message-ID: <20001212221753.A1526@localhost> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing Hi, I'm using OpenInventor 2.1 on Linux : OpenInventor rpms: sgi-OpenInventor-clients-2.1.5-6 sgi-OpenInventor-data-2.1.5-6 sgi-OpenInventor-devel-2.1.5-6 Linux: SuSE 7.0 (kernel 2.2.15) XFree 3.3.6 (xf86_3x-3.3.6-47.rpm) GLX (glx-000623-1.rpm) Mesa 3.2 (mesa-3.2-63.rpm) Hardware: PII 400 ATI 3D RagePro AGP (used as PCI by XF86 with driver Mach64) Note that I had to disable pixmap/font caches to be able to use glx. The text (SoText3) does not appear in my programs. Here is what I get: @@Created GLX Context.. invalid data type to GLX_data_size: 0x1407!! invalid data type to GLX_data_size: 0x1407!! invalid data type to GLX_data_size: 0x1407!! invalid data type to GLX_data_size: 0x1407!! I got the same error when I tried to redirect the display from a SGI running the same code to my Linux display. I guess there might be a bug in GLX, since the text works fine with Mesasoft, but this doesn't use the hardware acceleration of my card, which is sad. Any idea? Thanks. Erik Simon erik.simon@unine.ch From owner-info-inventor-dev@oss.sgi.com Wed Dec 13 22:06:38 2000 Received: by oss.sgi.com id ; Wed, 13 Dec 2000 22:06:28 -0800 Received: from pneumatic-tube.sgi.com ([204.94.214.22]:1554 "EHLO pneumatic-tube.sgi.com") by oss.sgi.com with ESMTP id ; Wed, 13 Dec 2000 22:06:09 -0800 Received: from boeing.engr.sgi.com (boeing.engr.sgi.com [130.62.55.185]) by pneumatic-tube.sgi.com (980327.SGI.8.8.8-aspam/980310.SGI-aspam) via ESMTP id WAA01925 for ; Wed, 13 Dec 2000 22:14:26 -0800 (PST) mail_from (flynnt@engr.sgi.com) Received: from localhost (flynnt@localhost) by boeing.engr.sgi.com (SGI-8.9.3/8.9.3) with ESMTP id WAA21616; Wed, 13 Dec 2000 22:04:31 -0800 (PST) X-Authentication-Warning: boeing.engr.sgi.com: flynnt owned process doing -bs Date: Wed, 13 Dec 2000 22:04:31 -0800 From: Tom Flynn To: Jonathan Lim cc: jamagallon@able.es, OpenInventor Devel List Subject: Re: [PATCH] GL12 Compatibility In-Reply-To: <200012130201.SAA89819@kinabalu.csd.sgi.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing For what it is worth, I haven't encountered such problems with Inventor as of yet, but have encountered such problem when porting Performer to Linux and had to make similar kinds of changes. Not all OpenGL vendors provide backwards compatability with older versions of OpenGL and it is not required of them. It may unfortunate for application and toolkit developers, but we have to work with it. -tom On Tue, 12 Dec 2000, Jonathan Lim wrote: > Hi all, > > I'm about the check the following changes in. Has anyone else encountered > similar problems and wish to add to the discussion? > > > Jonathan Lim _ Silicon Graphics _ Mountain View > GPS Graphics Computer Systems CA, USA > > > On Fri Dec 1 16:09:46 2000, jamagallon@able.es wrote: > > > > Hi everyone, > > > > I have been building OIV in a Mandrake 7.2 with XFree4.0.1. This xfree includes > > version 3.4 of Mesa, that is supposed to be GL1.2 compatible. That gives some > > drawbacks whe building Open Inventor. Basicly, inventor is coded against > > old OpenGL 1.0, where (for example) TexObjs were an extension, and recent > > GLs are 1.1 or 1.2, where TexObjs ar mandatory and functions do not have the > > EXT tail. > > > > Some implementations can support both interfaces (I think even mesa can do > > through preprocessor define flags), but I thin it is cleaner to write the > > code like: > > > > #ifndef GL_VERSION_1_1 > > glBindTextureEXT(GL_TEXTURE_2D, 0); > > #else > > glBindTexture(GL_TEXTURE_2D, 0); > > #endif > > > > I also think that for this kind of features (EXTs that have been adopted) > > one could skip the tests for extension presence, because the interface > > is supposed to be usable even if it does nothing (ie, it is just a > > software cache...), but I am not fully confident and so have made the > > minimal changes. If someone can tell me sure (hey, SGI people...) that > > this can be made, I can prepare the patch. > > > > Changes are: > > > > - A couple of casts to make g++2.96 shut up. > > - Poylgon offset interface. > > - Texture objects > > - Tesselator interface if GLU version is 1.2. > > > > Files affected are: > > apps/demos/textomatic/TextWrapper.c++ > > apps/nodes/Decal/Decal.c++ > > lib/database/src/so/elements/SoGLTextureImageElement.c++ > > lib/database/src/so/elements/SoGLCacheContextElement.c++ > > lib/database/src/so/nodes/SoShape.c++ > > lib/database/src/so/nodes/SoAsciiText.c++ > > > > Please, apply. I think it does not break anything and updates build > > compatibility. > > > > If you apply this on CVS, please let me or the list know, so I trash > > my patches and download corrected CVS. > > > > Thank you. > > > > -- > > Juan Antonio Magallon Lacarta #> cd /pub > > mailto:jamagallon@able.es #> more beer > -- "Mongooses are famous for their snake-fighting ability, and are almost always victorious because of their speed, agility, and timing and also because of their thick coat." From owner-info-inventor-dev@oss.sgi.com Thu Dec 14 01:20:39 2000 Received: by oss.sgi.com id ; Thu, 14 Dec 2000 01:20:19 -0800 Received: from deliverator.sgi.com ([204.94.214.10]:25890 "EHLO deliverator.sgi.com") by oss.sgi.com with ESMTP id ; Thu, 14 Dec 2000 01:20:01 -0800 Received: from boeing.engr.sgi.com (boeing.engr.sgi.com [130.62.55.185]) by deliverator.sgi.com (980309.SGI.8.8.8-aspam-6.2/980310.SGI-aspam) via ESMTP id BAA17693 for ; Thu, 14 Dec 2000 01:19:14 -0800 (PST) mail_from (flynnt@engr.sgi.com) Received: from localhost (flynnt@localhost) by boeing.engr.sgi.com (SGI-8.9.3/8.9.3) with ESMTP id BAA22882; Thu, 14 Dec 2000 01:18:27 -0800 (PST) X-Authentication-Warning: boeing.engr.sgi.com: flynnt owned process doing -bs Date: Thu, 14 Dec 2000 01:18:27 -0800 From: Tom Flynn To: Erik Simon cc: info-inventor-dev@oss.sgi.com Subject: Re: Text does not work in Linux (ATI/glx) In-Reply-To: <20001212221753.A1526@localhost> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing On Tue, 12 Dec 2000, Erik Simon wrote: > Hardware: > PII 400 > ATI 3D RagePro AGP (used as PCI by XF86 with driver Mach64) > > Note that I had to disable pixmap/font caches to be able to use glx. That's correct for the RagePro driver. > > The text (SoText3) does not appear in my programs. Here is what I get: > @@Created GLX Context.. > invalid data type to GLX_data_size: 0x1407!! > invalid data type to GLX_data_size: 0x1407!! > invalid data type to GLX_data_size: 0x1407!! > invalid data type to GLX_data_size: 0x1407!! Hmm, I'm not getting that error. What file are you trying to view? I'm currently writing from an ATI RagePro based Laptop using the utah-glx drivers. I'm using the 0.10-pre1 tarball and viewing /usr/share/data/models/simple/text3.iv > I got the same error when I tried to redirect the display from a SGI > running the same code to my Linux display. I guess there might be a > bug in GLX, since the text works fine with Mesasoft, but this doesn't > use the hardware acceleration of my card, which is sad. Well, given your description, it does sound like a bug in Utah-GLX (though can't say for sure without tracing down the error myself). First, if you haven't already, I'd try the 0.10-pre1 tarball. If that didn't work, I'd try compiling the utah-glx code with a more recent version of Mesa (say 3.2.1 or 3.4). If that still fails, it would help if a sample file (or code) that reproduces the problem was provided. Perhaps then, the problem could be isolated down to the OpenGL call(s) utah-glx is having a problem with. And then an appropriate bug could be filed with that project. hope that helps, tom > > Any idea? > > Thanks. > > Erik Simon > erik.simon@unine.ch > -- "Mongooses are famous for their snake-fighting ability, and are almost always victorious because of their speed, agility, and timing and also because of their thick coat." From owner-info-inventor-dev@oss.sgi.com Thu Dec 14 04:27:29 2000 Received: by oss.sgi.com id ; Thu, 14 Dec 2000 04:27:19 -0800 Received: from [195.158.3.33] ([195.158.3.33]:24846 "EHLO nuron.ccc.uz") by oss.sgi.com with ESMTP id ; Thu, 14 Dec 2000 04:26:57 -0800 Received: (from root@localhost) by nuron.ccc.uz (8.10.2/8.10.2) id eBECU0o32318; Thu, 14 Dec 2000 17:30:00 +0500 Received: from oss.sgi.com (IDENT:root@oss.sgi.com [216.32.174.190]) by nuron.ccc.uz (8.10.2/8.10.2) with ESMTP id eBCLLfH04836 for ; Wed, 13 Dec 2000 02:21:42 +0500 Received: by oss.sgi.com id ; Tue, 12 Dec 2000 13:17:26 -0800 Received: from client82-225.hispeed.ch ([62.2.82.225]:28933 "EHLO clochette.net2000.ch") by oss.sgi.com with ESMTP id ; Tue, 12 Dec 2000 13:16:53 -0800 Received: (from erik@localhost) by clochette.net2000.ch (8.10.2/8.10.2/SuSE Linux 8.10.0-0.3) id eBCLHrM01559 for info-inventor-dev@oss.sgi.com; Tue, 12 Dec 2000 22:17:53 +0100 Date: Tue, 12 Dec 2000 22:17:53 +0100 From: Erik Simon To: info-inventor-dev@oss.sgi.com Subject: Text does not work in Linux (ATI/glx) Message-ID: <20001212221753.A1526@localhost> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing Hi, I'm using OpenInventor 2.1 on Linux : OpenInventor rpms: sgi-OpenInventor-clients-2.1.5-6 sgi-OpenInventor-data-2.1.5-6 sgi-OpenInventor-devel-2.1.5-6 Linux: SuSE 7.0 (kernel 2.2.15) XFree 3.3.6 (xf86_3x-3.3.6-47.rpm) GLX (glx-000623-1.rpm) Mesa 3.2 (mesa-3.2-63.rpm) Hardware: PII 400 ATI 3D RagePro AGP (used as PCI by XF86 with driver Mach64) Note that I had to disable pixmap/font caches to be able to use glx. The text (SoText3) does not appear in my programs. Here is what I get: @@Created GLX Context.. invalid data type to GLX_data_size: 0x1407!! invalid data type to GLX_data_size: 0x1407!! invalid data type to GLX_data_size: 0x1407!! invalid data type to GLX_data_size: 0x1407!! I got the same error when I tried to redirect the display from a SGI running the same code to my Linux display. I guess there might be a bug in GLX, since the text works fine with Mesasoft, but this doesn't use the hardware acceleration of my card, which is sad. Any idea? Thanks. Erik Simon erik.simon@unine.ch From owner-info-inventor-dev@oss.sgi.com Thu Dec 14 04:29:59 2000 Received: by oss.sgi.com id ; Thu, 14 Dec 2000 04:29:49 -0800 Received: from [195.158.3.33] ([195.158.3.33]:30478 "EHLO nuron.ccc.uz") by oss.sgi.com with ESMTP id ; Thu, 14 Dec 2000 04:29:33 -0800 Received: (from root@localhost) by nuron.ccc.uz (8.10.2/8.10.2) id eBECWPG32372; Thu, 14 Dec 2000 17:32:25 +0500 Received: from oss.sgi.com (IDENT:root@oss.sgi.com [216.32.174.190]) by nuron.ccc.uz (8.10.2/8.10.2) with ESMTP id eBD27mH06380 for ; Wed, 13 Dec 2000 07:07:49 +0500 Received: by oss.sgi.com id ; Tue, 12 Dec 2000 18:03:39 -0800 Received: from pneumatic-tube.sgi.com ([204.94.214.22]:11887 "EHLO pneumatic-tube.sgi.com") by oss.sgi.com with ESMTP id ; Tue, 12 Dec 2000 18:03:27 -0800 Received: from kinabalu.csd.sgi.com (kinabalu.csd.sgi.com [150.166.1.88]) by pneumatic-tube.sgi.com (980327.SGI.8.8.8-aspam/980310.SGI-aspam) via ESMTP id SAA02924 for ; Tue, 12 Dec 2000 18:11:41 -0800 (PST) mail_from (jlim@kinabalu.csd.sgi.com) Received: by kinabalu.csd.sgi.com (980427.SGI.8.8.8/911001.SGI) id SAA89819; Tue, 12 Dec 2000 18:01:10 -0800 (PST) From: jlim@kinabalu.csd.sgi.com (Jonathan Lim) Message-Id: <200012130201.SAA89819@kinabalu.csd.sgi.com> Subject: Re: [PATCH] GL12 Compatibility To: jamagallon@able.es Date: Tue, 12 Dec 2000 18:01:10 -0800 (PST) Cc: info-inventor-dev@oss.sgi.com (OpenInventor Devel List) In-Reply-To: <20001202010946.I3982@werewolf.able.es> from "J . A . Magallon" at Dec 02, 2000 01:09:46 AM X-Mailer: ELM [version 2.5 PL2] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing Hi all, I'm about the check the following changes in. Has anyone else encountered similar problems and wish to add to the discussion? Jonathan Lim _ Silicon Graphics _ Mountain View GPS Graphics Computer Systems CA, USA On Fri Dec 1 16:09:46 2000, jamagallon@able.es wrote: > > Hi everyone, > > I have been building OIV in a Mandrake 7.2 with XFree4.0.1. This xfree includes > version 3.4 of Mesa, that is supposed to be GL1.2 compatible. That gives some > drawbacks whe building Open Inventor. Basicly, inventor is coded against > old OpenGL 1.0, where (for example) TexObjs were an extension, and recent > GLs are 1.1 or 1.2, where TexObjs ar mandatory and functions do not have the > EXT tail. > > Some implementations can support both interfaces (I think even mesa can do > through preprocessor define flags), but I thin it is cleaner to write the > code like: > > #ifndef GL_VERSION_1_1 > glBindTextureEXT(GL_TEXTURE_2D, 0); > #else > glBindTexture(GL_TEXTURE_2D, 0); > #endif > > I also think that for this kind of features (EXTs that have been adopted) > one could skip the tests for extension presence, because the interface > is supposed to be usable even if it does nothing (ie, it is just a > software cache...), but I am not fully confident and so have made the > minimal changes. If someone can tell me sure (hey, SGI people...) that > this can be made, I can prepare the patch. > > Changes are: > > - A couple of casts to make g++2.96 shut up. > - Poylgon offset interface. > - Texture objects > - Tesselator interface if GLU version is 1.2. > > Files affected are: > apps/demos/textomatic/TextWrapper.c++ > apps/nodes/Decal/Decal.c++ > lib/database/src/so/elements/SoGLTextureImageElement.c++ > lib/database/src/so/elements/SoGLCacheContextElement.c++ > lib/database/src/so/nodes/SoShape.c++ > lib/database/src/so/nodes/SoAsciiText.c++ > > Please, apply. I think it does not break anything and updates build > compatibility. > > If you apply this on CVS, please let me or the list know, so I trash > my patches and download corrected CVS. > > Thank you. > > -- > Juan Antonio Magallon Lacarta #> cd /pub > mailto:jamagallon@able.es #> more beer From owner-info-inventor-dev@oss.sgi.com Thu Dec 14 16:23:06 2000 Received: by oss.sgi.com id ; Thu, 14 Dec 2000 16:22:47 -0800 Received: from client82-58.hispeed.ch ([62.2.82.58]:40196 "EHLO clochette.net2000.ch") by oss.sgi.com with ESMTP id ; Thu, 14 Dec 2000 16:22:20 -0800 Received: (from erik@localhost) by clochette.net2000.ch (8.10.2/8.10.2/SuSE Linux 8.10.0-0.3) id eBF0NDP17755; Fri, 15 Dec 2000 01:23:13 +0100 Date: Fri, 15 Dec 2000 01:23:13 +0100 From: Erik Simon To: Tom Flynn Cc: info-inventor-dev@oss.sgi.com Subject: Re: Text does not work in Linux (ATI/glx) Message-ID: <20001215012313.A17223@clochette> References: <20001212221753.A1526@localhost> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: ; from flynnt@engr.sgi.com on Thu, Dec 14, 2000 at 01:18:27AM -0800 Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing On Thu, Dec 14, 2000 at 01:18:27AM -0800, Tom Flynn wrote: > On Tue, 12 Dec 2000, Erik Simon wrote: > > > Hardware: > > PII 400 > > ATI 3D RagePro AGP (used as PCI by XF86 with driver Mach64) > > > > Note that I had to disable pixmap/font caches to be able to use glx. > > That's correct for the RagePro driver. > > > > > The text (SoText3) does not appear in my programs. Here is what I get: > > @@Created GLX Context.. > > invalid data type to GLX_data_size: 0x1407!! > > invalid data type to GLX_data_size: 0x1407!! > > invalid data type to GLX_data_size: 0x1407!! > > invalid data type to GLX_data_size: 0x1407!! > > Hmm, I'm not getting that error. What file are you trying to view? I'm > currently writing from an ATI RagePro based Laptop using the utah-glx > drivers. I'm using the 0.10-pre1 tarball and viewing > /usr/share/data/models/simple/text3.iv > > > I got the same error when I tried to redirect the display from a SGI > > running the same code to my Linux display. I guess there might be a > > bug in GLX, since the text works fine with Mesasoft, but this doesn't > > use the hardware acceleration of my card, which is sad. > > Well, given your description, it does sound like a bug in Utah-GLX (though > can't say for sure without tracing down the error myself). > > First, if you haven't already, I'd try the 0.10-pre1 tarball. > If that didn't work, I'd try compiling the utah-glx code with a more > recent version of Mesa (say 3.2.1 or 3.4). If that still fails, it would > help if a sample file (or code) that reproduces the problem was provided. > Perhaps then, the problem could be isolated down to the OpenGL call(s) > utah-glx is having a problem with. And then an appropriate bug could be > filed with that project. > > hope that helps, > tom > > > > > Any idea? > > > > Thanks. > > > > Erik Simon > > erik.simon@unine.ch > > > > -- > "Mongooses are famous for their snake-fighting ability, and are > almost always victorious because of their speed, agility, and timing > and also because of their thick coat." Thanks a lot! For those who might have the same problem, I compiled Mesa 3.4 and the 0.10-pre1 tarball as Tom told me and it works! By the way, I was just trying to view a "hello world" written from scratch on a SGI. I didn't patch/recompile ivview yet and the version included in the rpms doesn't work on my system, but it has already been identified and solved I think. Regards. Erik From owner-info-inventor-dev@oss.sgi.com Sat Dec 16 07:02:45 2000 Received: by oss.sgi.com id ; Sat, 16 Dec 2000 07:02:26 -0800 Received: from bw150zhb.bluewin.ch ([195.186.1.68]:31407 "EHLO bw150zhb.bluewin.ch") by oss.sgi.com with ESMTP id ; Sat, 16 Dec 2000 07:02:10 -0800 Received: from localhost (gem167pub85.bluewin.ch [195.186.167.85]) by bw150zhb.bluewin.ch (8.9.3/8.9.3) with SMTP id QAA05980; Sat, 16 Dec 2000 16:01:48 +0100 (MET) Message-ID: <000501c06771$08a0dfa0$0100007f@localhost> From: "Erik Simon" To: Cc: Subject: Re: Text does not work in Linux (ATI/glx) Date: Sat, 16 Dec 2000 16:01:01 +0100 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2014.211 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2014.211 Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing Hi, Well, it doesn't work... I saw a nice text, right, but it was only because I compiled Mesa 3.4 and installed it, and the library was replaced. So the problem remains. I tried with the very last version of utah-glx from cvs and the very last 3.2dev version of mesa, also from cvs (without compiling mesa this time, but only glx against mesa), and it doesn't work. Note: glx doesn't compile (yet) with mesa 3.4, it was just a mistake, I compiled it against mesa 3.2. I'm not at home right now, so I cannot send a piece of code right now. Anyway, it doesn't work with any example from the Inventor Mentor that contains text (like the nice earth with Europe and Asia). Everything except text is fine. And everything works with the mesa library, but without acceleration. Shall I try to compile OpenInventor from the tarballs instead of using the rpms? I'll investigate the problem on the glx's mailing lists too. Regards. Erik From owner-info-inventor-dev@oss.sgi.com Wed Dec 20 06:31:01 2000 Received: by oss.sgi.com id ; Wed, 20 Dec 2000 06:30:51 -0800 Received: from ha1.rdc2.nsw.optushome.com.au ([203.164.2.50]:15843 "EHLO mss.rdc2.nsw.optushome.com.au") by oss.sgi.com with ESMTP id ; Wed, 20 Dec 2000 06:30:37 -0800 Received: from nigels.com ([203.164.64.206]) by mss.rdc2.nsw.optushome.com.au (InterMail vM.4.01.03.00 201-229-121) with ESMTP id <20001220143027.ZMJT28304.mss.rdc2.nsw.optushome.com.au@nigels.com>; Thu, 21 Dec 2000 01:30:27 +1100 Message-ID: <3A40C17A.750F1E40@nigels.com> Date: Thu, 21 Dec 2000 01:26:02 +1100 From: Nigel Stewart Reply-To: nigels@nigels.com Organization: The Cube Project X-Mailer: Mozilla 4.73 [en] (WinNT; U) X-Accept-Language: en,ja MIME-Version: 1.0 To: Inventor Developer List CC: Gerd Hesina Subject: Expressions of interest: GLUT Inventor Adaptor for Win32/Linux Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-info-inventor-dev@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;info-inventor-dev-outgoing I've been tinkering away at a C++ adaptor for GLUT and Inventor as part of the GLT (OpenGL Tools) library which is GPL and currently supports Win32 and Linux. The problem is that Christmas is fast approaching and it doesn't seem that I will have time to keep fiddling in Linux (Madrake 7.2) to get Inventor to compile so I can do a release of GLT with this new feature. If there is anyone here with a working Linux/Inventor installation and is willing to do some compiling, testing and trouble-shooting, please email me at nigels@nigels.com I don't believe that there will be any major issues, but I just don't have the opportunity to try the latest CVS snapshot on Mandrake to do this for myself. I'm also keen to support other Unixes or operating systems. For a demo of Inventor examples on Win32 including ivclock, maze and some Inventor Mentor examples, have a look at: http://www.nigels.com/glt/gltInventorDemo0.3.tgz These are precompiled Windows EXE and DLL. Mouse, keyboard and time events are all implemented - perhaps not to perfection, but certainly to a useful extent. They depend only on INVENTOR.DLL and GLUT32.DLL, which are included in the archive. Just repeating the point of this project - to combine the power and flexibility of Inventor with the portability and simplicity of GLUT. (Well, also to provide a nice and clean object-oriented C++ wrapper for GLUT) The scope of Open Inventor integration is all scenegraph functionality upto the SoSceneManager level. (ie. No OS-specific user interface components, etc) For more information, please refer to: http://www.nigels.com/glt/ Cheers, Nigel Stewart winiv.h ----------------------------------------------------------- #ifndef GLUTM_WINIV_H #define GLUTM_WINIV_H /* Glt OpenGL C++ Toolkit (Glt) Copyright (C) 2000 Nigel Stewart Email: nigels@nigels.com WWW: http://www.nigels.com/glt/ This library is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version. This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details. You should have received a copy of the GNU Lesser General Public License along with this library; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ #include #include #ifdef GLUTM_OPEN_INVENTOR #include class GlutWindowInventor : public GlutWindow, public SoSceneManager { friend GlutMaster; public: GlutWindowInventor( const std::string &title = titleDefault, int width = widthDefault, int height = heightDefault, int x = xDefault, int y = yDefault, unsigned int displayMode = displayModeDefault ); ~GlutWindowInventor(); void OnDisplay(); void OnIdle(); void OnReshape(int w, int h); void OnKeyboard(unsigned char key, int x, int y); void OnSpecial(int key, int x, int y); void OnMouse(int button, int state, int x, int y); void OnMotion(int x, int y); void OnPassiveMotion(int x, int y); protected: void OnPreTimer(unsigned char val); private: static void initOpenInventor(); static void inventorTick(); static bool _initOpenInventor; }; #endif #endif -- Nigel Stewart (nigels@nigels.com) Research Student, Software Developer Just two stops down the line at Zoo Station.