From sylvestre.ledru at scilab.org Mon Sep 7 17:07:35 2009 From: sylvestre.ledru at scilab.org (Sylvestre Ledru) Date: Mon, 07 Sep 2009 17:07:35 +0200 Subject: Massive upgrade of the build tools under Linux/Mac OS X & Unix Message-ID: <1252336055.4211.62147.camel@korcula.inria.fr> Hello, I just pushed a massive upgrade of the autotools used in the Scilab build system. * The most important point is the update of libtool (1.5.26 to 2.2.6a) asked by Jonathan. As said in the bug #4279, it greatly increases the compilation speed. Thanks for pointing that out. I should have done this sooner. For example, on my computer, I get $ time (./configure && make) [...] real 12m10.433s user 9m25.235s sys 1m44.595s * automake has been updated to version 1.11 * autoconf has been updated to version 2.64 * I also removed from Scilab source tree a embedded macro from autoconf (m4/libs_by_debian.m4). It was a modified file from autoconf by Debian which was fixing some autoconf bugs. Fixes which has been applied upstream since then and it was starting to cause some issues. I am aware that these versions may not yet available under Ubuntu (jaunty has the right version of libtool) and some other distributions. I am sorry for the problem it can causes but I had to push them now in order to have some time to fix the new bugs produced by libtool-bug-factory. As a workaround, you can: * remove the option --enable-maintainer-mode * ask me to rebuild {configure,Makefile.in} when needed * retrieve the packages from http://packages.ubuntu.com/jaunty/libtool (etc) Please, don't hesitate to report any bugs or undesired changes to: http://bugzilla.scilab.org/ Sylvestre From a.vandre.g at gmail.com Mon Sep 14 19:20:51 2009 From: a.vandre.g at gmail.com (=?ISO-8859-1?Q?Antonio_Vandr=E9?=) Date: Mon, 14 Sep 2009 14:20:51 -0300 Subject: Polynomy with multiple variables Message-ID: <9c89a4eb0909141020n4df3551di77d93905ac78d66e@mail.gmail.com> Hi friends. Polynomy with multiple variables can be implemented ? Always when i can try this error appear: !--error 144 Undefined operation for the given operands. check or define function %p_a_p for overloading. From raphael.langella at steria.cnes.fr Wed Sep 16 11:37:04 2009 From: raphael.langella at steria.cnes.fr (Langella Raphael) Date: Wed, 16 Sep 2009 11:37:04 +0200 Subject: [scilab-Users] scilab 5.1.1 compilation under Solaris 10 In-Reply-To: <1253031429.8710.47987.camel@korcula.inria.fr> Message-ID: <092785B790DCD043BA45401EDA43D9B504BB2DA7@cst-xch-003.cnesnet.ad.cnes.fr> > -----Message d'origine----- > First, would you mind switching this discussion on the > developer mailing list ? ( dev at lists.scilab.org ) Sure > Sorry, I missed that. > It is failing also in nwni or nw ? > If yes, could you try: > strace -f -o LogFileForSylvester.log ./bin/scilab -nwni -e "exit()" > If not > strace -f -o LogFileForSylvester.log ./bin/scilab -e "exit()" > and send me the log file With nwni, it seems to work, so I made an strace without the option. Here is the log. I killed it after a some time because it seemed to be stuck in a loop (and the file was already 4MB) >> But using mesa didn't solved the problem under Solaris 9, I still get : >> An error occurred when Scilab tried to open a Canvas: glXGetConfig >> failed: error code GLX_NO_EXTENSION > You should see if the glx is correctly configured/set on your Solaris. I thought the whole point of using mesa was to have complete software rendering. But I still need an X server with GLX support? I don't get it. -------------- next part -------------- A non-text attachment was scrubbed... Name: LogFileForSylvester.log.gz Type: application/x-gzip Size: 297142 bytes Desc: LogFileForSylvester.log.gz URL: From raphael.langella at steria.cnes.fr Wed Sep 16 11:53:03 2009 From: raphael.langella at steria.cnes.fr (Langella Raphael) Date: Wed, 16 Sep 2009 11:53:03 +0200 Subject: scilab 5.1.1 compilation under Solaris 10 Message-ID: <092785B790DCD043BA45401EDA43D9B504BB2DAC@cst-xch-003.cnesnet.ad.cnes.fr> > First, would you mind switching this discussion on the developer > mailing list ? ( dev at lists.scilab.org ) Sure > Sorry, I missed that. > It is failing also in nwni or nw ? > If yes, could you try: > strace -f -o LogFileForSylvester.log ./bin/scilab -nwni -e "exit()" > If not > strace -f -o LogFileForSylvester.log ./bin/scilab -e "exit()" > and send me the log file With nwni, it seems to work, so I made an strace without the option. Here is the log. I killed it after a some time because it seemed to be stuck in a loop (and the file was already 4MB) >> But using mesa didn't solved the problem under Solaris 9, I still get : >> An error occurred when Scilab tried to open a Canvas: glXGetConfig >> failed: error code GLX_NO_EXTENSION > You should see if the glx is correctly configured/set on your Solaris. I thought the whole point of using mesa was to have complete software rendering. But I still need an X server with GLX support? I don't get it. -------------- next part -------------- A non-text attachment was scrubbed... Name: LogFileForSylvester.log.gz Type: application/x-gzip Size: 297142 bytes Desc: LogFileForSylvester.log.gz URL: From sylvestre.ledru at scilab.org Wed Sep 16 13:51:30 2009 From: sylvestre.ledru at scilab.org (Sylvestre Ledru) Date: Wed, 16 Sep 2009 13:51:30 +0200 Subject: [Scilab-Dev] scilab 5.1.1 compilation under Solaris 10 In-Reply-To: <092785B790DCD043BA45401EDA43D9B504BB2DAC@cst-xch-003.cnesnet.ad.cnes.fr> References: <092785B790DCD043BA45401EDA43D9B504BB2DAC@cst-xch-003.cnesnet.ad.cnes.fr> Message-ID: <1253101890.8710.48046.camel@korcula.inria.fr> Le mercredi 16 septembre 2009 ? 11:53 +0200, Langella Raphael a ?crit : > > First, would you mind switching this discussion on the developer > > mailing list ? ( dev at lists.scilab.org ) > > Sure > > > Sorry, I missed that. > > It is failing also in nwni or nw ? > > If yes, could you try: > > strace -f -o LogFileForSylvester.log ./bin/scilab -nwni -e "exit()" > > If not > > strace -f -o LogFileForSylvester.log ./bin/scilab -e "exit()" > > and send me the log file > > With nwni, it seems to work, so I made an strace without the > option. Here is the log. I killed it after a some time > because it seemed to be stuck in a loop (and the file was already 4MB) Are you sure about the loop ? Usually, a strace just to open and close the application is about 60 mo. > >> But using mesa didn't solved the problem under Solaris 9, I still get > : > >> An error occurred when Scilab tried to open a Canvas: glXGetConfig > >> failed: error code GLX_NO_EXTENSION > > You should see if the glx is correctly configured/set on your Solaris. > > I thought the whole point of using mesa was to have complete > software rendering. But I still need an X server with GLX > support? I don't get it. I am no expert here but I think that mesa should provide GLX. Maybe the version of mesa you are using is not providing it! Sylvestre -- Sylvestre Ledru ------------------------- The Scilab Consortium Digiteo Domaine de Voluceau Rocquencourt - B.P. 105 78153 Le Chesnay Cedex France From raphael.langella at steria.cnes.fr Thu Sep 17 12:40:26 2009 From: raphael.langella at steria.cnes.fr (Langella Raphael) Date: Thu, 17 Sep 2009 12:40:26 +0200 Subject: [scilab-Users] scilab 5.1.1 compilation under Solaris 10 In-Reply-To: <1253005180.8710.41405.camel@korcula.inria.fr> Message-ID: <092785B790DCD043BA45401EDA43D9B504BB2E3E@cst-xch-003.cnesnet.ad.cnes.fr> > > I've also found out what was wrong with the make install > under Solaris 9. In libs/blas/Makefile, line 890, there is : > > > > for specialExt in $(MACROSSPECIALEXT); do \ > > > > But the MACROSSPECIALEXT is empty and the shell doesn't > like it. So I've delete this line and the next 3 ones, and > did the same in all the (~70!) other Makefiles. Now make > install works, but it really is an ugly workaround, there > should be a better way of fixing this. > :) > Indeed, there is easier way. > This stuff is declared in Makefile.incl.am and all makefile > can be regenerated with autoreconf. > You will find the patch attached to this email. > Apply it and launch autoreconf > it should fix you problem well, it should have, but it didn't: /bin/bash: -c: line 1: syntax error near unexpected token `;' /bin/bash: -c: line 1: `echo "-------- Install macros (if any) --------"; for dir in macros/ ; do /bin/bash /Produits/tmp/rla/Sol9/scilab-5.1.1/config/install-sh -d /Produits/publics/sparc.SunOS.5.9/scilab/5.1.1/share/scilab/modules/blas /$dir && if test -d ./$dir/; then FILELIST="./$dir/*.sci ./$dir/*.bin ./$dir/*.sce ./$dir/names ./$dir/lib"; if test -n ""; then for specialExt in ; do specialExtDir="$specialExtDir ./$dir/$specialExt"; done; FILELIST="$FILELIST $specialExtDir"; fi; for file in `ls -1 $FILELIST 2>/dev/null`; do echo "/Produits/publics/sparc.SunOS.5.9/bin/install -c -m 644 $file /Produits/publics/sparc.SunOS.5.9/scilab/5.1.1/share/scilab/modules/blas /$dir" ; /Produits/publics/sparc.SunOS.5.9/bin/install -c -m 644 "$file" /Produits/publics/sparc.SunOS.5.9/scilab/5.1.1/share/scilab/modules/blas /$dir ; done; fi; done' gmake[3]: *** [install-data-local] Error 2 gmake[3]: Leaving directory `/nfs/Produits/tmp/rla/Sol9/scilab-5.1.1/libs/blas' But I found another workaround. I changed: for specialExt in $(MACROSSPECIALEXT); do \ by for specialExt in "$(MACROSSPECIALEXT) " ; do \ It just didn't like the empty string, but with a space it's ok. by the way, my (stupid) shell version is: GNU bash, version 2.05.0(1)-release (sparc-sun-solaris2.9) From sylvestre.ledru at scilab.org Sat Sep 19 17:51:35 2009 From: sylvestre.ledru at scilab.org (Sylvestre Ledru) Date: Sat, 19 Sep 2009 17:51:35 +0200 Subject: SEP#34: Change of the main header file of call_scilab Message-ID: <1253375495.5485.4556.camel@zlarin> Hi, Here is a small SEP which has been already implemented. See bug http://bugzilla.scilab.org/show_bug.cgi?id=4795 Sylvestre -- ------------------------- Sylvestre Ledru ------------------------- The Scilab Consortium Digiteo Domaine de Voluceau Rocquencourt - B.P. 105 78153 Le Chesnay Cedex France -------------- next part -------------- A non-text attachment was scrubbed... Name: SEP_34_call_scilab_header_renamed.pdf Type: application/pdf Size: 187340 bytes Desc: not available URL: From sylvestre.ledru at scilab.org Tue Sep 22 09:18:09 2009 From: sylvestre.ledru at scilab.org (Sylvestre Ledru) Date: Tue, 22 Sep 2009 09:18:09 +0200 Subject: [Scilab-Dev] RE: [scilab-Users] scilab 5.1.1 compilation under Solaris 10 In-Reply-To: <092785B790DCD043BA45401EDA43D9B504BB2E3E@cst-xch-003.cnesnet.ad.cnes.fr> References: <092785B790DCD043BA45401EDA43D9B504BB2E3E@cst-xch-003.cnesnet.ad.cnes.fr> Message-ID: <1253603889.14112.7934.camel@korcula.inria.fr> Le jeudi 17 septembre 2009 ? 12:40 +0200, Langella Raphael a ?crit : > > > I've also found out what was wrong with the make install > > under Solaris 9. In libs/blas/Makefile, line 890, there is : > > > > > > for specialExt in $(MACROSSPECIALEXT); do \ > > > > > > But the MACROSSPECIALEXT is empty and the shell doesn't > > like it. So I've delete this line and the next 3 ones, and > > did the same in all the (~70!) other Makefiles. Now make > > install works, but it really is an ugly workaround, there > > should be a better way of fixing this. > > :) > > Indeed, there is easier way. > > This stuff is declared in Makefile.incl.am and all makefile > > can be regenerated with autoreconf. > > You will find the patch attached to this email. > > Apply it and launch autoreconf > > it should fix you problem > > well, it should have, but it didn't: > > /bin/bash: -c: line 1: syntax error near unexpected token `;' > /bin/bash: -c: line 1: `echo "-------- Install macros (if any) > --------"; for dir in macros/ ; do /bin/bash > /Produits/tmp/rla/Sol9/scilab-5.1.1/config/install-sh -d > /Produits/publics/sparc.SunOS.5.9/scilab/5.1.1/share/scilab/modules/blas > /$dir && if test -d ./$dir/; then FILELIST="./$dir/*.sci ./$dir/*.bin > ./$dir/*.sce ./$dir/names ./$dir/lib"; if test -n ""; then for > specialExt in ; do specialExtDir="$specialExtDir ./$dir/$specialExt"; > done; FILELIST="$FILELIST $specialExtDir"; fi; for file in `ls -1 > $FILELIST 2>/dev/null`; do echo > "/Produits/publics/sparc.SunOS.5.9/bin/install -c -m 644 $file > /Produits/publics/sparc.SunOS.5.9/scilab/5.1.1/share/scilab/modules/blas > /$dir" ; /Produits/publics/sparc.SunOS.5.9/bin/install -c -m 644 > "$file" > /Produits/publics/sparc.SunOS.5.9/scilab/5.1.1/share/scilab/modules/blas > /$dir ; done; fi; done' > gmake[3]: *** [install-data-local] Error 2 > gmake[3]: Leaving directory > `/nfs/Produits/tmp/rla/Sol9/scilab-5.1.1/libs/blas' > > But I found another workaround. I changed: > for specialExt in $(MACROSSPECIALEXT); do \ > by > for specialExt in "$(MACROSSPECIALEXT) " ; do \ > > It just didn't like the empty string, but with a space it's ok. I pushed a few for that. I added a test: if test -n "$(MACROSSPECIALEXT)"; then This should do it > by the way, my (stupid) shell version is: > GNU bash, version 2.05.0(1)-release (sparc-sun-solaris2.9) $ bash -version GNU bash, version 4.0.28(1)-release (i486-pc-linux-gnu) Sylvestre -- Sylvestre Ledru ------------------------- The Scilab Consortium Digiteo Domaine de Voluceau Rocquencourt - B.P. 105 78153 Le Chesnay Cedex France From raphael.langella at steria.cnes.fr Tue Sep 22 10:11:26 2009 From: raphael.langella at steria.cnes.fr (Langella Raphael) Date: Tue, 22 Sep 2009 10:11:26 +0200 Subject: [Scilab-Dev] RE: [scilab-Users] scilab 5.1.1 compilation under Solaris 10 In-Reply-To: <1253603889.14112.7934.camel@korcula.inria.fr> Message-ID: <092785B790DCD043BA45401EDA43D9B504BB2F8E@cst-xch-003.cnesnet.ad.cnes.fr> > -----Message d'origine----- > De : Sylvestre Ledru [mailto:sylvestre.ledru at scilab.org] > Envoy? : mardi 22 septembre 2009 09:18 > ? : dev at lists.scilab.org > Objet : Re: [Scilab-Dev] RE: [scilab-Users] scilab 5.1.1 > compilation under Solaris 10 > > But I found another workaround. I changed: > > for specialExt in $(MACROSSPECIALEXT); do \ by for specialExt in > > "$(MACROSSPECIALEXT) " ; do \ > > > > It just didn't like the empty string, but with a space it's ok. > I pushed a few for that. > I added a test: > if test -n "$(MACROSSPECIALEXT)"; then > > This should do it I tried your patch, I saw that you added the test and I understand that it should fix it but it didn't. It still gave me: syntax error near unexpected token `;' I needed both your fix and mine so that the make install really works. My Makefile.incl.am looks like this: if test -n "$(MACROSSPECIALEXT)"; then \ for specialExt in "$(MACROSSPECIALEXT) "; do \ specialExtDir="$$specialExtDir $(srcdir)/$$dir/$$specialExt"; \ done; \ FILELIST="$$FILELIST $$specialExtDir"; \ fi; \ From Lukas.Malec at vso-praha.eu Thu Sep 24 19:00:35 2009 From: Lukas.Malec at vso-praha.eu (=?iso-8859-2?Q?Malec_Luk=E1=B9?=) Date: Thu, 24 Sep 2009 19:00:35 +0200 Subject: installation of some toolboxes Message-ID: <2952C79E4A4C8A4396B17CA34694B819A8789AD6FA@dc.vso-praha.eu> Good day, please to consult me with the following installation problem of some toolboxes. The installation on MS Windows XP Professional proceed, but during the compillation process with the builder script exec builder.sce, the message emerge, that a Fortran or C compiler is requested. During the command exec loader.sce invoke the message similar to past: -->exec loader.sce -->// Generated by builder.sce - Do Not Edit -->mode(-1); link(libs + libsext); !--error 236 link: The shared archive was not loaded: The filename, directory name, or volume label syntax is incorrect. at line 21 of exec file called by : exec loader.sce I don?t know, where to draw down the Microsoft C compiler, if it is necessary, and how to install it. Please, don?t you know the answer to this problem? The answer will help me very much, some toolboxes are what I had been searching for. Thank you and Best regards, Luk?? -------------- next part -------------- An HTML attachment was scrubbed... URL: From sylvestre.ledru at scilab.org Tue Sep 29 17:15:17 2009 From: sylvestre.ledru at scilab.org (Sylvestre Ledru) Date: Tue, 29 Sep 2009 17:15:17 +0200 Subject: SEP #35: Selection of the installed headers Message-ID: <1254237317.14112.76845.camel@korcula.inria.fr> Hello, Just a new SEP which selects the installed headers. For now, in the Scilab install, there are way too many header files. This SEP proposes to leave only the essential/mandatory ones. Message to module/toolbox authors: Please note that I might have forgot a few in this list. Please drop an email to let me know. I will update the list. This SEP, among others, in an other step in the goal of a clear, easy, consistent and clean API to interact with Scilab. Sylvestre -- Sylvestre Ledru ------------------------- The Scilab Consortium Digiteo Domaine de Voluceau Rocquencourt - B.P. 105 78153 Le Chesnay Cedex France -------------- next part -------------- A non-text attachment was scrubbed... Name: SEP_35_selection_of_installed_header.pdf Type: application/pdf Size: 474496 bytes Desc: not available URL: