Math-intensive work in 1.4.x impossible.
I installed the rpm for LyX 1.4.2 distributed in the Fedora 5 extras repository. I've found that it's still impossible to work intensively on mathematical expressions with LyX 1.4.x. After I perform a relatively small amount of editing, inside and outside math mode, the window simply crashes and then I have to recover the document. That happened to me 3 times in about 10 minutes of editing... The same problem occurred with all versions 1.4.1 and 1.4.2 that I tested. Every time I have to come back to version 1.3.7, which is fortunately very robust. I would like to post a debug report on this. How can I channel the ouput of lyx -dbg any into a file? lyx -dbg any lyx_bug.txt did not work... -- Rudi Gaelzer Department of Physics Institute of Physics and Mathematics Federal University of Pelotas BRAZIL Registered linux user # 153741
Re: Math-intensive work in 1.4.x impossible.
I installed the rpm for LyX 1.4.2 distributed in the Fedora 5 extras repository. I've found that it's still impossible to work intensively on mathematical expressions with LyX 1.4.x. After I perform a relatively small amount of editing, inside and outside math mode, the window simply crashes and then I have to recover the document. That happened to me 3 times in about 10 minutes of editing... The same problem occurred with all versions 1.4.1 and 1.4.2 that I tested. Every time I have to come back to version 1.3.7, which is fortunately very robust. I would like to post a debug report on this. How can I channel the ouput of lyx -dbg any into a file? lyx -dbg any lyx_bug.txt did not work... Try lyx -dbg any 2 lyx_bug.txt Sorry, I can't help you on the other stuff, I don't keep track of LyX develpment at the moment :(. Cheers, Michael -- Rudi Gaelzer Department of Physics Institute of Physics and Mathematics Federal University of Pelotas BRAZIL Registered linux user # 153741
Re: Math-intensive work in 1.4.x impossible.
Rudi == Rudi Gaelzer [EMAIL PROTECTED] writes: Rudi I've found that it's still impossible to work intensively on Rudi mathematical expressions with LyX 1.4.x. After I perform a Rudi relatively small amount of editing, inside and outside math Rudi mode, the window simply crashes and then I have to recover the Rudi document. That happened to me 3 times in about 10 minutes of Rudi editing... Try to provide us with repeatable crashes. I am currently working to fix some crashes related to sub/superscripts, for example. JMarc
Re: Math-intensive work in 1.4.x impossible.
Jean-Marc, I don't think this is possible. I mean, the crash seems to occur at any instant. I can not identify any particular sequence of keystrokes that will ignite it... the only pattern I can see is that the window crashes always after a lot of intensive editing in mathed mode, usually when I'm leaving or entering a math environment. I could send the bug report obtained from lyx -dbg any 2 lyx_bug.txt, as Michael Abshoff suggested, but the file, even when compressed with bzip2, is above the 6b limit of the mailer-deamon. Should I create a bug report at bugzilla? On Thursday 17 August 2006 09:51, Jean-Marc Lasgouttes wrote: Rudi == Rudi Gaelzer [EMAIL PROTECTED] writes: Rudi I've found that it's still impossible to work intensively on Rudi mathematical expressions with LyX 1.4.x. After I perform a Rudi relatively small amount of editing, inside and outside math Rudi mode, the window simply crashes and then I have to recover the Rudi document. That happened to me 3 times in about 10 minutes of Rudi editing... Try to provide us with repeatable crashes. I am currently working to fix some crashes related to sub/superscripts, for example. JMarc -- Rudi Gaelzer Departamento de Física Instituto de Física e Matemática Fundação Universidade Federal de Pelotas Caixa Postal 354 - Campus UFPel 96010-900 Pelotas - RS Fone: +55-53-275-7416 FAX: +55-53-275-7343 Usuário de Linux Registrado Nº 153741
Re: Math-intensive work in 1.4.x impossible.
Rudi Gaelzer wrote: Jean-Marc, I don't think this is possible. I mean, the crash seems to occur at any instant. I can not identify any particular sequence of keystrokes that will ignite it... the only pattern I can see is that the window crashes always after a lot of intensive editing in mathed mode, usually when I'm leaving or entering a math environment. Is it related to using instant preview? Jürgen
Re: Math-intensive work in 1.4.x impossible.
Am Donnerstag, 17. August 2006 18:27 schrieb Rudi Gaelzer: Jean-Marc, I don't think this is possible. I mean, the crash seems to occur at any instant. I can not identify any particular sequence of keystrokes that will ignite it... the only pattern I can see is that the window crashes always after a lot of intensive editing in mathed mode, usually when I'm leaving or entering a math environment. Did you compile LyX yourself? If yes, it would be nice if you could compile it with debugging symbols (run configure with --enable-debug). Then run LyX in gdb, and when it crashes press bt for a backtrace and send that. That might give us useful information. I could send the bug report obtained from lyx -dbg any 2 lyx_bug.txt, as Michael Abshoff suggested, but the file, even when compressed with bzip2, is above the 6b limit of the mailer-deamon. You could try lyx -dbg mathed,action 2 lyx_bug.txt instead. If that is still too big then simply send the last 1000 lines. Should I create a bug report at bugzilla? You can, but it does not help much if the bug is not reproducible and we don't have a backtrace and/or debug output. If you manage to get a backtrace or debug output please do. The strange thing is that heavy math editing works rather well for me in 1.4.3svn. What OS/distribution and what compiler do you use? We have reported crashes with gcc 4.1 that might be related to this, see e.g. http://bugzilla.lyx.org/show_bug.cgi?id=2677 Georg
Re: Math-intensive work in 1.4.x impossible.
Georg Baum wrote: The strange thing is that heavy math editing works rather well for me in 1.4.3svn. What OS/distribution and what compiler do you use? We have reported crashes with gcc 4.1 that might be related to this, see e.g. http://bugzilla.lyx.org/show_bug.cgi?id=2677 fc5/gcc-4.1, so that may be it -- Rex
Re: Math-intensive work in 1.4.x impossible.
Indeed, I'm running lyx on FC5 with gcc 4.1.x. Attached is the bug report obtained with lyx -dbg mathed,action 2 lyx_bug.txt Apparently, the crash really occurs when I exit mathed mode and instant preview should kick in... Could this be the bug related to gcc? Thanks. On Thursday 17 August 2006 14:17, Georg Baum wrote: Am Donnerstag, 17. August 2006 18:27 schrieb Rudi Gaelzer: Jean-Marc, I don't think this is possible. I mean, the crash seems to occur at any instant. I can not identify any particular sequence of keystrokes that will ignite it... the only pattern I can see is that the window crashes always after a lot of intensive editing in mathed mode, usually when I'm leaving or entering a math environment. Did you compile LyX yourself? If yes, it would be nice if you could compile it with debugging symbols (run configure with --enable-debug). Then run LyX in gdb, and when it crashes press bt for a backtrace and send that. That might give us useful information. I could send the bug report obtained from lyx -dbg any 2 lyx_bug.txt, as Michael Abshoff suggested, but the file, even when compressed with bzip2, is above the 6b limit of the mailer-deamon. You could try lyx -dbg mathed,action 2 lyx_bug.txt instead. If that is still too big then simply send the last 1000 lines. Should I create a bug report at bugzilla? You can, but it does not help much if the bug is not reproducible and we don't have a backtrace and/or debug output. If you manage to get a backtrace or debug output please do. The strange thing is that heavy math editing works rather well for me in 1.4.3svn. What OS/distribution and what compiler do you use? We have reported crashes with gcc 4.1 that might be related to this, see e.g. http://bugzilla.lyx.org/show_bug.cgi?id=2677 Georg -- Rudi Gaelzer Departamento de Física Instituto de Física e Matemática Fundação Universidade Federal de Pelotas Caixa Postal 354 - Campus UFPel 96010-900 Pelotas - RS Fone: +55-53-275-7416 FAX: +55-53-275-7343 Usuário de Linux Registrado Nº 153741 lyx_bug.txt.bz2 Description: BZip2 compressed data
Re: Math-intensive work in 1.4.x impossible.
On Thursday 17 August 2006 18:56, Rudi Gaelzer wrote: Indeed, I'm running lyx on FC5 with gcc 4.1.x. Attached is the bug report obtained with lyx -dbg mathed,action 2 lyx_bug.txt Apparently, the crash really occurs when I exit mathed mode and instant preview should kick in... Could this be the bug related to gcc? Thanks. When writing my thesis, where I had hundreds of math expressions I had to turn off instant preview or else I would see the same behaviour. I am running FC5 as well but as reported in this thread this seems to be an issue with gcc 4.1 that is not exclusive of FC5. :-( -- José Abílio
Re: Math-intensive work in 1.4.x impossible.
Am Donnerstag, 17. August 2006 20:00 schrieb José Matos: I am running FC5 as well but as reported in this thread this seems to be an issue with gcc 4.1 that is not exclusive of FC5. :-( Yes. Fortunately we know the cause and will fix it for 1.4.3: The included boost 1.32 does not work well with gcc 4.1. This is fixed in boost 1.33. Rudi, I think it is best if you stick with 1.3.7 until we have fixed this issue and then try again. Georg
Re: Math-intensive work in 1.4.x impossible.
On Thursday 17 August 2006 19:21, Georg Baum wrote: Rudi, I think it is best if you stick with 1.3.7 until we have fixed this issue and then try again. Or turn off instant preview for the moment. :-) Georg -- José Abílio
Re: Math-intensive work in 1.4.x impossible.
Rudi Gaelzer wrote: Apparently, the crash really occurs when I exit mathed mode and instant preview should kick in... Could this be the bug related to gcc? Yes. It apparently happens with the boost version included to LyX 1.4 and gcc 4.1. See this bug: http://bugzilla.lyx.org/show_bug.cgi?id=2677 and this one: http://bugzilla.lyx.org/show_bug.cgi?id=2662 Your best bet for now is to turn preview-latex off. Jürgen
Re: Math-intensive work in 1.4.x impossible.
Georg == Georg Baum [EMAIL PROTECTED] writes: Georg Am Donnerstag, 17. August 2006 20:00 schrieb José Matos: I am running FC5 as well but as reported in this thread this seems to Georg be an issue with gcc 4.1 that is not exclusive of FC5. :-( Georg Yes. Fortunately we know the cause and will fix it for 1.4.3: Georg The included boost 1.32 does not work well with gcc 4.1. This Georg is fixed in boost 1.33. What do we have to do to upgrade boost like that? JMarc
Math-intensive work in 1.4.x impossible.
I installed the rpm for LyX 1.4.2 distributed in the Fedora 5 extras repository. I've found that it's still impossible to work intensively on mathematical expressions with LyX 1.4.x. After I perform a relatively small amount of editing, inside and outside math mode, the window simply crashes and then I have to recover the document. That happened to me 3 times in about 10 minutes of editing... The same problem occurred with all versions 1.4.1 and 1.4.2 that I tested. Every time I have to come back to version 1.3.7, which is fortunately very robust. I would like to post a debug report on this. How can I channel the ouput of lyx -dbg any into a file? lyx -dbg any lyx_bug.txt did not work... -- Rudi Gaelzer Department of Physics Institute of Physics and Mathematics Federal University of Pelotas BRAZIL Registered linux user # 153741
Re: Math-intensive work in 1.4.x impossible.
I installed the rpm for LyX 1.4.2 distributed in the Fedora 5 extras repository. I've found that it's still impossible to work intensively on mathematical expressions with LyX 1.4.x. After I perform a relatively small amount of editing, inside and outside math mode, the window simply crashes and then I have to recover the document. That happened to me 3 times in about 10 minutes of editing... The same problem occurred with all versions 1.4.1 and 1.4.2 that I tested. Every time I have to come back to version 1.3.7, which is fortunately very robust. I would like to post a debug report on this. How can I channel the ouput of lyx -dbg any into a file? lyx -dbg any lyx_bug.txt did not work... Try lyx -dbg any 2 lyx_bug.txt Sorry, I can't help you on the other stuff, I don't keep track of LyX develpment at the moment :(. Cheers, Michael -- Rudi Gaelzer Department of Physics Institute of Physics and Mathematics Federal University of Pelotas BRAZIL Registered linux user # 153741
Re: Math-intensive work in 1.4.x impossible.
Rudi == Rudi Gaelzer [EMAIL PROTECTED] writes: Rudi I've found that it's still impossible to work intensively on Rudi mathematical expressions with LyX 1.4.x. After I perform a Rudi relatively small amount of editing, inside and outside math Rudi mode, the window simply crashes and then I have to recover the Rudi document. That happened to me 3 times in about 10 minutes of Rudi editing... Try to provide us with repeatable crashes. I am currently working to fix some crashes related to sub/superscripts, for example. JMarc
Re: Math-intensive work in 1.4.x impossible.
Jean-Marc, I don't think this is possible. I mean, the crash seems to occur at any instant. I can not identify any particular sequence of keystrokes that will ignite it... the only pattern I can see is that the window crashes always after a lot of intensive editing in mathed mode, usually when I'm leaving or entering a math environment. I could send the bug report obtained from lyx -dbg any 2 lyx_bug.txt, as Michael Abshoff suggested, but the file, even when compressed with bzip2, is above the 6b limit of the mailer-deamon. Should I create a bug report at bugzilla? On Thursday 17 August 2006 09:51, Jean-Marc Lasgouttes wrote: Rudi == Rudi Gaelzer [EMAIL PROTECTED] writes: Rudi I've found that it's still impossible to work intensively on Rudi mathematical expressions with LyX 1.4.x. After I perform a Rudi relatively small amount of editing, inside and outside math Rudi mode, the window simply crashes and then I have to recover the Rudi document. That happened to me 3 times in about 10 minutes of Rudi editing... Try to provide us with repeatable crashes. I am currently working to fix some crashes related to sub/superscripts, for example. JMarc -- Rudi Gaelzer Departamento de Física Instituto de Física e Matemática Fundação Universidade Federal de Pelotas Caixa Postal 354 - Campus UFPel 96010-900 Pelotas - RS Fone: +55-53-275-7416 FAX: +55-53-275-7343 Usuário de Linux Registrado Nº 153741
Re: Math-intensive work in 1.4.x impossible.
Rudi Gaelzer wrote: Jean-Marc, I don't think this is possible. I mean, the crash seems to occur at any instant. I can not identify any particular sequence of keystrokes that will ignite it... the only pattern I can see is that the window crashes always after a lot of intensive editing in mathed mode, usually when I'm leaving or entering a math environment. Is it related to using instant preview? Jürgen
Re: Math-intensive work in 1.4.x impossible.
Am Donnerstag, 17. August 2006 18:27 schrieb Rudi Gaelzer: Jean-Marc, I don't think this is possible. I mean, the crash seems to occur at any instant. I can not identify any particular sequence of keystrokes that will ignite it... the only pattern I can see is that the window crashes always after a lot of intensive editing in mathed mode, usually when I'm leaving or entering a math environment. Did you compile LyX yourself? If yes, it would be nice if you could compile it with debugging symbols (run configure with --enable-debug). Then run LyX in gdb, and when it crashes press bt for a backtrace and send that. That might give us useful information. I could send the bug report obtained from lyx -dbg any 2 lyx_bug.txt, as Michael Abshoff suggested, but the file, even when compressed with bzip2, is above the 6b limit of the mailer-deamon. You could try lyx -dbg mathed,action 2 lyx_bug.txt instead. If that is still too big then simply send the last 1000 lines. Should I create a bug report at bugzilla? You can, but it does not help much if the bug is not reproducible and we don't have a backtrace and/or debug output. If you manage to get a backtrace or debug output please do. The strange thing is that heavy math editing works rather well for me in 1.4.3svn. What OS/distribution and what compiler do you use? We have reported crashes with gcc 4.1 that might be related to this, see e.g. http://bugzilla.lyx.org/show_bug.cgi?id=2677 Georg
Re: Math-intensive work in 1.4.x impossible.
Georg Baum wrote: The strange thing is that heavy math editing works rather well for me in 1.4.3svn. What OS/distribution and what compiler do you use? We have reported crashes with gcc 4.1 that might be related to this, see e.g. http://bugzilla.lyx.org/show_bug.cgi?id=2677 fc5/gcc-4.1, so that may be it -- Rex
Re: Math-intensive work in 1.4.x impossible.
Indeed, I'm running lyx on FC5 with gcc 4.1.x. Attached is the bug report obtained with lyx -dbg mathed,action 2 lyx_bug.txt Apparently, the crash really occurs when I exit mathed mode and instant preview should kick in... Could this be the bug related to gcc? Thanks. On Thursday 17 August 2006 14:17, Georg Baum wrote: Am Donnerstag, 17. August 2006 18:27 schrieb Rudi Gaelzer: Jean-Marc, I don't think this is possible. I mean, the crash seems to occur at any instant. I can not identify any particular sequence of keystrokes that will ignite it... the only pattern I can see is that the window crashes always after a lot of intensive editing in mathed mode, usually when I'm leaving or entering a math environment. Did you compile LyX yourself? If yes, it would be nice if you could compile it with debugging symbols (run configure with --enable-debug). Then run LyX in gdb, and when it crashes press bt for a backtrace and send that. That might give us useful information. I could send the bug report obtained from lyx -dbg any 2 lyx_bug.txt, as Michael Abshoff suggested, but the file, even when compressed with bzip2, is above the 6b limit of the mailer-deamon. You could try lyx -dbg mathed,action 2 lyx_bug.txt instead. If that is still too big then simply send the last 1000 lines. Should I create a bug report at bugzilla? You can, but it does not help much if the bug is not reproducible and we don't have a backtrace and/or debug output. If you manage to get a backtrace or debug output please do. The strange thing is that heavy math editing works rather well for me in 1.4.3svn. What OS/distribution and what compiler do you use? We have reported crashes with gcc 4.1 that might be related to this, see e.g. http://bugzilla.lyx.org/show_bug.cgi?id=2677 Georg -- Rudi Gaelzer Departamento de Física Instituto de Física e Matemática Fundação Universidade Federal de Pelotas Caixa Postal 354 - Campus UFPel 96010-900 Pelotas - RS Fone: +55-53-275-7416 FAX: +55-53-275-7343 Usuário de Linux Registrado Nº 153741 lyx_bug.txt.bz2 Description: BZip2 compressed data
Re: Math-intensive work in 1.4.x impossible.
On Thursday 17 August 2006 18:56, Rudi Gaelzer wrote: Indeed, I'm running lyx on FC5 with gcc 4.1.x. Attached is the bug report obtained with lyx -dbg mathed,action 2 lyx_bug.txt Apparently, the crash really occurs when I exit mathed mode and instant preview should kick in... Could this be the bug related to gcc? Thanks. When writing my thesis, where I had hundreds of math expressions I had to turn off instant preview or else I would see the same behaviour. I am running FC5 as well but as reported in this thread this seems to be an issue with gcc 4.1 that is not exclusive of FC5. :-( -- José Abílio
Re: Math-intensive work in 1.4.x impossible.
Am Donnerstag, 17. August 2006 20:00 schrieb José Matos: I am running FC5 as well but as reported in this thread this seems to be an issue with gcc 4.1 that is not exclusive of FC5. :-( Yes. Fortunately we know the cause and will fix it for 1.4.3: The included boost 1.32 does not work well with gcc 4.1. This is fixed in boost 1.33. Rudi, I think it is best if you stick with 1.3.7 until we have fixed this issue and then try again. Georg
Re: Math-intensive work in 1.4.x impossible.
On Thursday 17 August 2006 19:21, Georg Baum wrote: Rudi, I think it is best if you stick with 1.3.7 until we have fixed this issue and then try again. Or turn off instant preview for the moment. :-) Georg -- José Abílio
Re: Math-intensive work in 1.4.x impossible.
Rudi Gaelzer wrote: Apparently, the crash really occurs when I exit mathed mode and instant preview should kick in... Could this be the bug related to gcc? Yes. It apparently happens with the boost version included to LyX 1.4 and gcc 4.1. See this bug: http://bugzilla.lyx.org/show_bug.cgi?id=2677 and this one: http://bugzilla.lyx.org/show_bug.cgi?id=2662 Your best bet for now is to turn preview-latex off. Jürgen
Re: Math-intensive work in 1.4.x impossible.
Georg == Georg Baum [EMAIL PROTECTED] writes: Georg Am Donnerstag, 17. August 2006 20:00 schrieb José Matos: I am running FC5 as well but as reported in this thread this seems to Georg be an issue with gcc 4.1 that is not exclusive of FC5. :-( Georg Yes. Fortunately we know the cause and will fix it for 1.4.3: Georg The included boost 1.32 does not work well with gcc 4.1. This Georg is fixed in boost 1.33. What do we have to do to upgrade boost like that? JMarc
Math-intensive work in 1.4.x impossible.
I installed the rpm for LyX 1.4.2 distributed in the Fedora 5 extras repository. I've found that it's still impossible to work intensively on mathematical expressions with LyX 1.4.x. After I perform a relatively small amount of editing, inside and outside math mode, the window simply crashes and then I have to recover the document. That happened to me 3 times in about 10 minutes of editing... The same problem occurred with all versions 1.4.1 and 1.4.2 that I tested. Every time I have to come back to version 1.3.7, which is fortunately very robust. I would like to post a debug report on this. How can I channel the ouput of "lyx -dbg any" into a file? "lyx -dbg any > lyx_bug.txt" did not work... -- Rudi Gaelzer Department of Physics Institute of Physics and Mathematics Federal University of Pelotas BRAZIL Registered linux user # 153741
Re: Math-intensive work in 1.4.x impossible.
> I installed the rpm for LyX 1.4.2 distributed in the Fedora 5 extras > repository. > > I've found that it's still impossible to work intensively on mathematical > expressions with LyX 1.4.x. After I perform a relatively small amount of > editing, inside and outside math mode, the window simply crashes and then > I > have to recover the document. That happened to me 3 times in about 10 > minutes of editing... > > The same problem occurred with all versions 1.4.1 and 1.4.2 that I tested. > Every time I have to come back to version 1.3.7, which is fortunately very > robust. > > I would like to post a debug report on this. How can I channel the ouput > of "lyx -dbg any" into a file? "lyx -dbg any > lyx_bug.txt" did not > work... Try "lyx -dbg any 2> lyx_bug.txt &" Sorry, I can't help you on the other stuff, I don't keep track of LyX develpment at the moment :(. Cheers, Michael > -- > Rudi Gaelzer > Department of Physics > Institute of Physics and Mathematics > Federal University of Pelotas > BRAZIL > Registered linux user # 153741 >
Re: Math-intensive work in 1.4.x impossible.
> "Rudi" == Rudi Gaelzer <[EMAIL PROTECTED]> writes: Rudi> I've found that it's still impossible to work intensively on Rudi> mathematical expressions with LyX 1.4.x. After I perform a Rudi> relatively small amount of editing, inside and outside math Rudi> mode, the window simply crashes and then I have to recover the Rudi> document. That happened to me 3 times in about 10 minutes of Rudi> editing... Try to provide us with repeatable crashes. I am currently working to fix some crashes related to sub/superscripts, for example. JMarc
Re: Math-intensive work in 1.4.x impossible.
Jean-Marc, I don't think this is possible. I mean, the crash seems to occur at any instant. I can not identify any particular sequence of keystrokes that will ignite it... the only pattern I can see is that the window crashes always after a lot of intensive editing in mathed mode, usually when I'm leaving or entering a math environment. I could send the bug report obtained from "lyx -dbg any 2> lyx_bug.txt", as Michael Abshoff suggested, but the file, even when compressed with bzip2, is above the 6b limit of the mailer-deamon. Should I create a bug report at bugzilla? On Thursday 17 August 2006 09:51, Jean-Marc Lasgouttes wrote: > > "Rudi" == Rudi Gaelzer <[EMAIL PROTECTED]> writes: > > Rudi> I've found that it's still impossible to work intensively on > Rudi> mathematical expressions with LyX 1.4.x. After I perform a > Rudi> relatively small amount of editing, inside and outside math > Rudi> mode, the window simply crashes and then I have to recover the > Rudi> document. That happened to me 3 times in about 10 minutes of > Rudi> editing... > > Try to provide us with repeatable crashes. I am currently working to > fix some crashes related to sub/superscripts, for example. > > JMarc -- Rudi Gaelzer Departamento de Física Instituto de Física e Matemática Fundação Universidade Federal de Pelotas Caixa Postal 354 - Campus UFPel 96010-900 Pelotas - RS Fone: +55-53-275-7416 FAX: +55-53-275-7343 Usuário de Linux Registrado Nº 153741
Re: Math-intensive work in 1.4.x impossible.
Rudi Gaelzer wrote: > Jean-Marc, > I don't think this is possible. I mean, the crash seems to occur at any > instant. I can not identify any particular sequence of keystrokes that > will ignite it... the only pattern I can see is that the window crashes > always after a lot of intensive editing in mathed mode, usually when I'm > leaving or entering a math environment. Is it related to using instant preview? Jürgen
Re: Math-intensive work in 1.4.x impossible.
Am Donnerstag, 17. August 2006 18:27 schrieb Rudi Gaelzer: > Jean-Marc, > I don't think this is possible. I mean, the crash seems to occur at any > instant. I can not identify any particular sequence of keystrokes that will > ignite it... the only pattern I can see is that the window crashes always > after a lot of intensive editing in mathed mode, usually when I'm leaving or > entering a math environment. Did you compile LyX yourself? If yes, it would be nice if you could compile it with debugging symbols (run configure with --enable-debug). Then run LyX in gdb, and when it crashes press bt for a backtrace and send that. That might give us useful information. > I could send the bug report obtained from "lyx -dbg any 2> lyx_bug.txt", as > Michael Abshoff suggested, but the file, even when compressed with bzip2, is > above the 6b limit of the mailer-deamon. You could try lyx -dbg mathed,action 2> lyx_bug.txt instead. If that is still too big then simply send the last 1000 lines. > Should I create a bug report at bugzilla? You can, but it does not help much if the bug is not reproducible and we don't have a backtrace and/or debug output. If you manage to get a backtrace or debug output please do. The strange thing is that heavy math editing works rather well for me in 1.4.3svn. What OS/distribution and what compiler do you use? We have reported crashes with gcc 4.1 that might be related to this, see e.g. http://bugzilla.lyx.org/show_bug.cgi?id=2677 Georg
Re: Math-intensive work in 1.4.x impossible.
Georg Baum wrote: > The strange thing is that heavy math editing works rather well for me in > 1.4.3svn. What OS/distribution and what compiler do you use? We have > reported crashes with gcc 4.1 that might be related to this, see e.g. > http://bugzilla.lyx.org/show_bug.cgi?id=2677 fc5/gcc-4.1, so that may be it -- Rex
Re: Math-intensive work in 1.4.x impossible.
Indeed, I'm running lyx on FC5 with gcc 4.1.x. Attached is the bug report obtained with "lyx -dbg mathed,action 2> lyx_bug.txt" Apparently, the crash really occurs when I exit mathed mode and instant preview should kick in... Could this be the bug related to gcc? Thanks. On Thursday 17 August 2006 14:17, Georg Baum wrote: > Am Donnerstag, 17. August 2006 18:27 schrieb Rudi Gaelzer: > > Jean-Marc, > > I don't think this is possible. I mean, the crash seems to occur at any > > instant. I can not identify any particular sequence of keystrokes that > > will > > > ignite it... the only pattern I can see is that the window crashes > > always > > > after a lot of intensive editing in mathed mode, usually when I'm leaving > > or > > > entering a math environment. > > Did you compile LyX yourself? If yes, it would be nice if you could compile > it with debugging symbols (run configure with --enable-debug). Then run > LyX in gdb, and when it crashes press bt for a backtrace and send that. > That might give us useful information. > > > I could send the bug report obtained from "lyx -dbg any 2> lyx_bug.txt", > > as > > > Michael Abshoff suggested, but the file, even when compressed with bzip2, > > is > > > above the 6b limit of the mailer-deamon. > > You could try > > lyx -dbg mathed,action 2> lyx_bug.txt > > instead. If that is still too big then simply send the last 1000 lines. > > > Should I create a bug report at bugzilla? > > You can, but it does not help much if the bug is not reproducible and we > don't have a backtrace and/or debug output. If you manage to get a > backtrace or debug output please do. > > The strange thing is that heavy math editing works rather well for me in > 1.4.3svn. What OS/distribution and what compiler do you use? We have > reported crashes with gcc 4.1 that might be related to this, see e.g. > http://bugzilla.lyx.org/show_bug.cgi?id=2677 > > > Georg -- Rudi Gaelzer Departamento de Física Instituto de Física e Matemática Fundação Universidade Federal de Pelotas Caixa Postal 354 - Campus UFPel 96010-900 Pelotas - RS Fone: +55-53-275-7416 FAX: +55-53-275-7343 Usuário de Linux Registrado Nº 153741 lyx_bug.txt.bz2 Description: BZip2 compressed data
Re: Math-intensive work in 1.4.x impossible.
On Thursday 17 August 2006 18:56, Rudi Gaelzer wrote: > Indeed, I'm running lyx on FC5 with gcc 4.1.x. > Attached is the bug report obtained with "lyx -dbg mathed,action 2> > lyx_bug.txt" > > Apparently, the crash really occurs when I exit mathed mode and instant > preview should kick in... > > Could this be the bug related to gcc? > > Thanks. When writing my thesis, where I had hundreds of math expressions I had to turn off instant preview or else I would see the same behaviour. I am running FC5 as well but as reported in this thread this seems to be an issue with gcc 4.1 that is not exclusive of FC5. :-( -- José Abílio
Re: Math-intensive work in 1.4.x impossible.
Am Donnerstag, 17. August 2006 20:00 schrieb José Matos: > I am running FC5 as well but as reported in this thread this seems to be an > issue with gcc 4.1 that is not exclusive of FC5. :-( Yes. Fortunately we know the cause and will fix it for 1.4.3: The included boost 1.32 does not work well with gcc 4.1. This is fixed in boost 1.33. Rudi, I think it is best if you stick with 1.3.7 until we have fixed this issue and then try again. Georg
Re: Math-intensive work in 1.4.x impossible.
On Thursday 17 August 2006 19:21, Georg Baum wrote: > Rudi, I think it is best if you stick with 1.3.7 until we have fixed this > issue and then try again. Or turn off instant preview for the moment. :-) > Georg -- José Abílio
Re: Math-intensive work in 1.4.x impossible.
Rudi Gaelzer wrote: > Apparently, the crash really occurs when I exit mathed mode and instant > preview should kick in... > > Could this be the bug related to gcc? Yes. It apparently happens with the boost version included to LyX 1.4 and gcc 4.1. See this bug: http://bugzilla.lyx.org/show_bug.cgi?id=2677 and this one: http://bugzilla.lyx.org/show_bug.cgi?id=2662 Your best bet for now is to turn preview-latex off. Jürgen
Re: Math-intensive work in 1.4.x impossible.
> "Georg" == Georg Baum <[EMAIL PROTECTED]> writes: Georg> Am Donnerstag, 17. August 2006 20:00 schrieb José Matos: >> I am running FC5 as well but as reported in this thread this seems >> to Georg> be an >> issue with gcc 4.1 that is not exclusive of FC5. :-( Georg> Yes. Fortunately we know the cause and will fix it for 1.4.3: Georg> The included boost 1.32 does not work well with gcc 4.1. This Georg> is fixed in boost 1.33. What do we have to do to upgrade boost like that? JMarc
