Обсуждение: Debian Alpha (unstable) build produces bad binary
Hi!
I built and installed Debian packages for pgadmin3 and pgadmin3-libwxgtk2ud2.5
following the instructions on the pgadmin3 web site. When I run pgadmin3,
however, it segfaults right away. Using strace, the last few lines look like
this:
stat("/usr/local/share/locale/fileutils.mo", 0x11ffff3e0) = -1 ENOENT (No such
file or directory)
stat("./en/LC_MESSAGES/fileutils.mo", 0x11ffff3e0) = -1 ENOENT (No such file
or directory)
stat("./en/fileutils.mo", 0x11ffff3e0) = -1 ENOENT (No such file or
directory)
stat("./fileutils.mo", 0x11ffff3e0) = -1 ENOENT (No such file or
directory)
--- SIGSEGV (Segmentation fault) @ 0 (0) ---
+++ killed by SIGSEGV +++
Using strace on my Mandrake i586 box, however, the same lines are produced in
the strace output (except the segfault, natch). Even though the fileutils.mo
file is not found, the program runs just fine on the Mandrake box.
I don't know if there's something missing from the Debian package build or
what. Anybody have any suggestions on what I should look for?
Regards,
.... Bob
Bob Kimble wrote: >I don't know if there's something missing from the Debian package build or >what. Anybody have any suggestions on what I should look for? > > > Could you also send a backtrace from gdb at the time of the segfault? Like this: user$ gdb pgAdmin3 (gdb) r SEGFAULT (gdb) bt Then cut-and-paste that output into an e-mail message (NOTE: the output from GDB will look different than this, but the commands to enter are the same). ahp
Bob Kimble wrote:
>Hi!
>
>I built and installed Debian packages for pgadmin3 and pgadmin3-libwxgtk2ud2.5
>following the instructions on the pgadmin3 web site. When I run pgadmin3,
>however, it segfaults right away. Using strace, the last few lines look like
>this:
>
>stat("/usr/local/share/locale/fileutils.mo", 0x11ffff3e0) = -1 ENOENT (No such
>file or directory)
>stat("./en/LC_MESSAGES/fileutils.mo", 0x11ffff3e0) = -1 ENOENT (No such file
>or directory)
>stat("./en/fileutils.mo", 0x11ffff3e0) = -1 ENOENT (No such file or
>directory)
>stat("./fileutils.mo", 0x11ffff3e0) = -1 ENOENT (No such file or
>directory)
>--- SIGSEGV (Segmentation fault) @ 0 (0) ---
>+++ killed by SIGSEGV +++
>
>
Hi Bob,
I don't believe that the SEGV is caused by the missing file, you can
check this by commenting out the load attempt at line 184 in pgAdmin3.cpp.
A stack backtrace would be very helpful to tell more.
Regards,
Andreas
Please ignore my earlier post -- I discovered the "c" command. Here's what I believe you want: (gdb) r Starting program: /usr/bin/pgadmin3 (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...[New Thread 16384 (LWP 28930)] (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...Cannot access memory at address 0x0 (gdb) c Continuing. (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)... Program received signal SIGSEGV, Segmentation fault. 0x000002000074f454 in g_type_is_a () from /usr/lib/libgobject-2.0.so.0 (gdb) bt #0 0x000002000074f454 in g_type_is_a () from /usr/lib/libgobject-2.0.so.0 #1 0x0000020000476028 in gtk_type_new () from /usr/lib/libgtk-x11-2.0.so.0 #2 0x00000001207fe5cc in gtk_pizza_new () #3 0x00000001205e8040 in wxTopLevelWindowGTK::Create(wxWindow*, int, wxString const&, wxPoint const&, wxSize const&, long, wxString const&) () #4 0x00000001206512c4 in wxDialog::Create(wxWindow*, int, wxString const&, wxPoint const&, wxSize const&, long, wxString const&) () #5 0x000000012069cb24 in wxAnyChoiceDialog::Create(wxWindow*, wxString const&, wxString const&, int, wxString const*, long, wxPoint const&, long) () #6 0x000000012069d4f8 in wxSingleChoiceDialog::Create(wxWindow*, wxString const&, wxString const&, int, wxString const*, char**, long, wxPoint const&) () #7 0x000000012069d408 in wxSingleChoiceDialog::wxSingleChoiceDialog(wxWindow*, wxString const&, wxString const&, int, wxString const*, char**, long, wxPoint const&) () #8 0x000000012069c2dc in wxGetSingleChoiceIndex(wxString const&, wxString const&, int, wxString const*, wxWindow*, int, int, bool, int, int) () #9 0x000000012018a604 in pgAdmin3::OnInit() () #10 0x000000012018b7dc in wxAppConsole::CallOnInit() () #11 0x00000001205269b4 in wxEntry(int&, wchar_t**) () #12 0x0000000120526b04 in wxEntry(int&, char**) () #13 0x00000001201886f0 in main () (gdb) Regards, .... Bob
Here's the backtrace. Doesn't look like much help to me. Is there some way to get the Debian packages to build an executable with the debugging symbols? (gdb) r Starting program: /usr/bin/pgadmin3 (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...[New Thread 16384 (LWP 28899)] (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...Cannot access memory at address 0x0 (gdb) bt #0 0x0000000000000000 in ?? () (gdb)
Bob Kimble wrote: >It appears that I'm running version 2.2.4 of gtk (that's the version of the >libgtk2.0-0 package I have installed). I'm running Debian unstable on that >machine in order to play with the most current software. I don't really know >how to install a newer version. I use apt-get, dselect, and dpkg to install >everything. > 2.2.4 is already the newest. I'm using 2.2.1. > >BTW -- I can run pgadmin3 just fine on my other machines, so this is not a >critical issue for me. > Still, it's a bit unsatisfying. It should run on *all* machines, at least if they're Linux with recent gtk2. Try to put the line LanguageId=0 in the file ~/.pgadmin3, this will prevent pgAdmin3 from asking for the language. You may still change it later from the options dialog. Actually, I doubt that this will help a lot, but rather crash as soon you open another dialog, but let's give it a try. >I use PostgreSQL for my database back end, > Good taste! > and pgadmin3 is just brilliant -- it's a truly excellent piece of software. > Thanks a lot! Regards, Andreas
On Monday, Sep 15, 2003, at 12:52 US/Eastern, Bob Kimble wrote: > 0x000002000074f454 in g_type_is_a () from /usr/lib/libgobject-2.0.so.0 > (gdb) bt > #0 0x000002000074f454 in g_type_is_a () from > /usr/lib/libgobject-2.0.so.0 > #1 0x0000020000476028 in gtk_type_new () from > /usr/lib/libgtk-x11-2.0.so.0 > #2 0x00000001207fe5cc in gtk_pizza_new () > #3 0x00000001205e8040 in wxTopLevelWindowGTK::Create(wxWindow*, int, > wxString > const&, wxPoint const&, wxSize const&, long, wxString const&) () > #4 0x00000001206512c4 in wxDialog::Create(wxWindow*, int, wxString > const&, > wxPoint const&, wxSize const&, long, wxString const&) () > #5 0x000000012069cb24 in wxAnyChoiceDialog::Create(wxWindow*, wxString > const&, wxString const&, int, wxString const*, long, wxPoint const&, > long) () > #6 0x000000012069d4f8 in wxSingleChoiceDialog::Create(wxWindow*, > wxString > const&, wxString const&, int, wxString const*, char**, long, wxPoint > const&) > () > #7 0x000000012069d408 in > wxSingleChoiceDialog::wxSingleChoiceDialog(wxWindow*, wxString const&, > wxString const&, int, wxString const*, char**, long, wxPoint const&) () > #8 0x000000012069c2dc in wxGetSingleChoiceIndex(wxString const&, > wxString > const&, int, wxString const*, wxWindow*, int, int, bool, int, int) () > #9 0x000000012018a604 in pgAdmin3::OnInit() () > #10 0x000000012018b7dc in wxAppConsole::CallOnInit() () > #11 0x00000001205269b4 in wxEntry(int&, wchar_t**) () > #12 0x0000000120526b04 in wxEntry(int&, char**) () > #13 0x00000001201886f0 in main () > (gdb) > I can almost guarantee that one of the parameters to one of the functions on this stack trace is NULL. Every time I've seen one of these problems (and we've had a couple with pgAdmin3), the problem is almost always that. The best way to tell is to step your way through the program and print out the parameters before they get passed. I don't know how familiar you are with gdb, but maybe you can give this a shot. Or maybe you can give Andreas access to your server to look at it. :-) ahp
On Monday 15 September 2003 18:12, Andreas Pflug wrote: > Still, it's a bit unsatisfying. It should run on *all* machines, at > least if they're Linux with recent gtk2. > Try to put the line > LanguageId=0 > in the file ~/.pgadmin3, this will prevent pgAdmin3 from asking for the > language. You may still change it later from the options dialog. > Actually, I doubt that this will help a lot, but rather crash as soon > you open another dialog, but let's give it a try. Here's the new output of gdb: GNU gdb 5.3.90_2003-08-24-cvs-debian Copyright 2003 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type "show copying" to see the conditions. There is absolutely no warranty for GDB. Type "show warranty" for details. This GDB was configured as "alpha-linux"...(no debugging symbols found)... (gdb) r Starting program: /usr/bin/pgadmin3 (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)...[New Thread 16384 (LWP 31597)] (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...Cannot access memory at address 0x0 (gdb) c Continuing. (no debugging symbols found)...(no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)... Program received signal SIGSEGV, Segmentation fault. 0x000002000074f454 in g_type_is_a () from /usr/lib/libgobject-2.0.so.0 (gdb) bt #0 0x000002000074f454 in g_type_is_a () from /usr/lib/libgobject-2.0.so.0 #1 0x0000020000476028 in gtk_type_new () from /usr/lib/libgtk-x11-2.0.so.0 #2 0x00000001207fe5cc in gtk_pizza_new () #3 0x00000001205e8040 in wxTopLevelWindowGTK::Create(wxWindow*, int, wxString const&, wxPoint const&, wxSize const&, long, wxString const&) () #4 0x00000001206544e0 in wxFrame::Create(wxWindow*, int, wxString const&, wxPoint const&, wxSize const&, long, wxString const&) () #5 0x00000001202cf400 in frmSplash::frmSplash(wxFrame*) () #6 0x0000000120189390 in pgAdmin3::OnInit() () #7 0x000000012018b7dc in wxAppConsole::CallOnInit() () #8 0x00000001205269b4 in wxEntry(int&, wchar_t**) () #9 0x0000000120526b04 in wxEntry(int&, char**) () #10 0x00000001201886f0 in main () (gdb) Looks like it's about the same thing, only at a different spot in pgAdmin3::OnInit(), if I'm reading the output correctly. Is it possible that some bug was introduced into the 2.2.4 version of gtk that I installed? Is there another program I could run to test gtk? I have everything but the kitchen sink installed on that machine -- as I said, it's just for playing around with mostly. Regards, .... Bob
Hi bob, Andreas, I've had a look to my build environment yesterday evening. The debian/unstable packages are actually built against libgtk2.0-0 2.2.2-3. libgtk2 Debian packages have been upgraded in the last days...I'll have a try tonight with this new version and tell youif it runs well on i386 and may be hppa (it will depend on my free time). Andreas, couldn't it be a compiler issue too (I had many problems with gcc on alpha in the past) ? Regards, Raphaël. ----Message d'origine---- >De: Bob Kimble <bob@iplicity.com> >A: Andreas Pflug <pgadmin@pse-consulting.de> >Sujet: Re: [pgadmin-support] Debian Alpha (unstable) build produces bad binary >Date: Mon, 15 Sep 2003 16:10:12 -0400 >Copie à: pgadmin-support@postgresql.org > >On Monday 15 September 2003 13:06, Andreas Pflug wrote: >> Bob Kimble wrote: ><snip> >> Well that looks good, or really bad. It's wxGetSingleChoiceIndex in line >> 201 pgAdmin3.cpp, which has just 4 parameters (rest is default), and >> they're certainly ok. >> wxWindows tries to create a simple dialog, and it fails deep in gtk. >> Could you try a newer gtk? Please tell us which version you're using >> now, and which you're trying. Maybe that helps, because the wx code has >> been working in million of other cases. >> >> Regards, >> Andreas > >It appears that I'm running version 2.2.4 of gtk (that's the version of the >libgtk2.0-0 package I have installed). I'm running Debian unstable on that >machine in order to play with the most current software. I don't really know >how to install a newer version. I use apt-get, dselect, and dpkg to install >everything. > >BTW -- I can run pgadmin3 just fine on my other machines, so this is not a >critical issue for me. I'm just trying to do what I can to help with the >development effort. I'm not a C or C++ programmer -- I pretty much use just >Java and Python these days. I use PostgreSQL for my database back end, and >pgadmin3 is just brilliant -- it's a truly excellent piece of software. > >Regards, > >..... Bob > > >---------------------------(end of broadcast)--------------------------- >TIP 6: Have you searched our list archives? > > http://archives.postgresql.org > >
Bob Kimble wrote: > >Looks like it's about the same thing, only at a different spot in >pgAdmin3::OnInit(), if I'm reading the output correctly. > Yeah, looks as I expected. > >Is it possible that some bug was introduced into the 2.2.4 version of gtk that >I installed? > That's what I rate with quite a higher probability. >Is there another program I could run to test gtk? I have >everything but the kitchen sink installed on that machine -- as I said, it's >just for playing around with mostly. > You could try the wx samples. I bet they show the same problem. Regards, Andreas
blacknoz@club-internet.fr wrote: >Hi bob, Andreas, > >I've had a look to my build environment yesterday evening. >The debian/unstable packages are actually built against libgtk2.0-0 2.2.2-3. >libgtk2 Debian packages have been upgraded in the last days...I'll have a try tonight with this new version and tell youif it runs well on i386 and may be hppa (it will depend on my free time). > >Andreas, couldn't it be a compiler issue too (I had many problems with gcc on alpha in the past) ? > Could be... But there's no "dangerous" call that I would suspect a compiler could make a mistake. Regards, Andreas
On Monday 15 September 2003 19:11, Adam H. Pendleton wrote: > I can almost guarantee that one of the parameters to one of the > functions on this stack trace is NULL. Every time I've seen one of > these problems (and we've had a couple with pgAdmin3), the problem is > almost always that. The best way to tell is to step your way through > the program and print out the parameters before they get passed. I > don't know how familiar you are with gdb, but maybe you can give this a > shot. Or maybe you can give Andreas access to your server to look at > it. :-) I'll try to figure out how to print out the parameters as you indicate. However, if Andreas wants access to my machine, it's no problem. I'll be quite happy to give him an account. It's just an AlphaPC sitting in my basement at home. Regards, .... Bob
Hi all, I upgraded all packages of the debian/unstable environment except postgresql and did some tests with the "old" pgadmin3 beta2 package and also latest pgadmin3 snapshots and "libgtk2.0-0 2.2.4", it runs well on i386. I'm currently doing a new build of wxWin 2.5 with libgtk2 2.2.4, I'll will then rebuild pga3 and tell you if it's ok. For the moment it seems to be an issue specific to alpha. Regards, Raphaël Andreas Pflug wrote: > blacknoz@club-internet.fr wrote: > >> Hi bob, Andreas, >> >> I've had a look to my build environment yesterday evening. >> The debian/unstable packages are actually built against libgtk2.0-0 >> 2.2.2-3. >> libgtk2 Debian packages have been upgraded in the last days...I'll >> have a try tonight with this new version and tell you if it runs well >> on i386 and may be hppa (it will depend on my free time). >> >> Andreas, couldn't it be a compiler issue too (I had many problems >> with gcc on alpha in the past) ? >> > Could be... > But there's no "dangerous" call that I would suspect a compiler could > make a mistake. > > Regards, > Andreas > > > > ---------------------------(end of broadcast)--------------------------- > TIP 7: don't forget to increase your free space map settings >
Raphaël Enrici wrote: > Hi all, > I upgraded all packages of the debian/unstable environment except > postgresql and did some tests with the "old" pgadmin3 beta2 package > and also latest pgadmin3 snapshots and "libgtk2.0-0 2.2.4", it runs > well on i386. I'm currently doing a new build of wxWin 2.5 with > libgtk2 2.2.4, I'll will then rebuild pga3 and tell you if it's ok. > For the moment it seems to be an issue specific to alpha. Andreas, I did a new build of wx and pgadmin3 latest snapshots in an up to date unstable environment. It runs well on i386. Did you get time to have a look to this problem on alpha ? Regards, Raphaël > > > Regards, > Raphaël > > Andreas Pflug wrote: > >> blacknoz@club-internet.fr wrote: >> >>> Hi bob, Andreas, >>> >>> I've had a look to my build environment yesterday evening. >>> The debian/unstable packages are actually built against libgtk2.0-0 >>> 2.2.2-3. >>> libgtk2 Debian packages have been upgraded in the last days...I'll >>> have a try tonight with this new version and tell you if it runs >>> well on i386 and may be hppa (it will depend on my free time). >>> >>> Andreas, couldn't it be a compiler issue too (I had many problems >>> with gcc on alpha in the past) ? >>> >> Could be... >> But there's no "dangerous" call that I would suspect a compiler could >> make a mistake. >> >> Regards, >> Andreas >> >> >> >> ---------------------------(end of broadcast)--------------------------- >> TIP 7: don't forget to increase your free space map settings >> > > > > ---------------------------(end of broadcast)--------------------------- > TIP 2: you can get off all lists at once with the unregister command > (send "unregister YourEmailAddressHere" to majordomo@postgresql.org) >
Raphaël Enrici wrote: > Raphaël Enrici wrote: > >> Hi all, >> I upgraded all packages of the debian/unstable environment except >> postgresql and did some tests with the "old" pgadmin3 beta2 package >> and also latest pgadmin3 snapshots and "libgtk2.0-0 2.2.4", it runs >> well on i386. I'm currently doing a new build of wxWin 2.5 with >> libgtk2 2.2.4, I'll will then rebuild pga3 and tell you if it's ok. >> For the moment it seems to be an issue specific to alpha. > > > Andreas, > I did a new build of wx and pgadmin3 latest snapshots in an up to date > unstable environment. It runs well on i386. > Did you get time to have a look to this problem on alpha ? > No, I didn't. Actually, that call from wx into gtk is m_wxwindow = gtk_pizza_new(), so wx has little chance to screw up the parameters, there are none. Bob, did you try a wx sample? This can only be solved by somebody with deep gtk experience, I believe. Regards, Andreas
On Thursday 18 September 2003 04:24, Andreas Pflug wrote: > Raphaël Enrici wrote: > > Raphaël Enrici wrote: > >> Hi all, > >> I upgraded all packages of the debian/unstable environment except > >> postgresql and did some tests with the "old" pgadmin3 beta2 package > >> and also latest pgadmin3 snapshots and "libgtk2.0-0 2.2.4", it runs > >> well on i386. I'm currently doing a new build of wxWin 2.5 with > >> libgtk2 2.2.4, I'll will then rebuild pga3 and tell you if it's ok. > >> For the moment it seems to be an issue specific to alpha. > > > > Andreas, > > I did a new build of wx and pgadmin3 latest snapshots in an up to date > > unstable environment. It runs well on i386. > > Did you get time to have a look to this problem on alpha ? > > No, I didn't. > Actually, that call from wx into gtk is > m_wxwindow = gtk_pizza_new(), > so wx has little chance to screw up the parameters, there are none. > Bob, did you try a wx sample? This can only be solved by somebody with > deep gtk experience, I believe. > > Regards, > Andreas I'm not really sure what a good example of a wx program is. The only one that seems to be one that I can find on my system is wxgambitcl, which seems to work fine. If somebody can tell me a Debian package to install, I'll do so and run the programs it contains. Regards, .... Bob
Bob Kimble wrote: >I'm not really sure what a good example of a wx program is. > Hi Bob, the internat sample is quite good, it uses the same wxGetSingleChoiceIndex as pgAdmin3 to select the language. When running the program, please make sure that Japanese and Georgian are presented to select, this is a proof that the program was compiled correctly with wxUSE_UNICODE. Regards, Andreas
Bob Kimble wrote: >On Thursday 18 September 2003 10:06, Andreas Pflug wrote: > > >>Bob Kimble wrote: >> >> >>>I'm not really sure what a good example of a wx program is. >>> >>> >>Hi Bob, >> >>the internat sample is quite good, it uses the same >>wxGetSingleChoiceIndex as pgAdmin3 to select the language. When running >>the program, please make sure that Japanese and Georgian are presented >>to select, this is a proof that the program was compiled correctly with >>wxUSE_UNICODE. >> >>Regards, >>Andreas >> >> > >Hi Andreas, > >It took me a little while to figure out how to compile internat, mainly >because I don't seem to have a version of wx-config in my path. Not sure >where that's supposed to come from. Anyway, I found a version of wx-config >that seems to work, and I compiled internat, which breaks in exactly the same >way as pgadmin3. I suppose that means pgadmin3 is not the culprit. > Hi Bob, It would have been quite a surprise if that sample didn't crash too. Please post that issue at wxWindow's bug system. Regards, Andreas
Bob Kimble wrote: >On Thursday 18 September 2003 11:30, Andreas Pflug wrote: > > >>Hi Bob, >>It would have been quite a surprise if that sample didn't crash too. >>Please post that issue at wxWindow's bug system. >> >>Regards, >>Andreas >> >> > >Sounds good. However, the version of wxWindows used by the pgadmin3 package >comes from www.pgadmin.org. I suspect I'm going to have to wait a bit before >the whole thing will be ready for the Alpha. > Bob, can you check wxWindows CVS head on that Alpha machine? We will rarely update our wx package, so to get Alpha fixed the only way is to use an official wx version and have one of their developers work on that. It might even be a bug in gtk2, which would need an escalation to the gtk team... Regards, Andreas
On Thursday 18 September 2003 11:30, Andreas Pflug wrote: > Hi Bob, > It would have been quite a surprise if that sample didn't crash too. > Please post that issue at wxWindow's bug system. > > Regards, > Andreas Sounds good. However, the version of wxWindows used by the pgadmin3 package comes from www.pgadmin.org. I suspect I'm going to have to wait a bit before the whole thing will be ready for the Alpha. Thanks for your assistance. Regards, .... Bob
On Thursday 18 September 2003 10:06, Andreas Pflug wrote: > Bob Kimble wrote: > >I'm not really sure what a good example of a wx program is. > > Hi Bob, > > the internat sample is quite good, it uses the same > wxGetSingleChoiceIndex as pgAdmin3 to select the language. When running > the program, please make sure that Japanese and Georgian are presented > to select, this is a proof that the program was compiled correctly with > wxUSE_UNICODE. > > Regards, > Andreas Hi Andreas, It took me a little while to figure out how to compile internat, mainly because I don't seem to have a version of wx-config in my path. Not sure where that's supposed to come from. Anyway, I found a version of wx-config that seems to work, and I compiled internat, which breaks in exactly the same way as pgadmin3. I suppose that means pgadmin3 is not the culprit. Regards, .... Bob
On Thursday 18 September 2003 12:17, Andreas Pflug wrote: > Bob Kimble wrote: > >On Thursday 18 September 2003 11:30, Andreas Pflug wrote: > >>Hi Bob, > >>It would have been quite a surprise if that sample didn't crash too. > >>Please post that issue at wxWindow's bug system. > >> > >>Regards, > >>Andreas > > > >Sounds good. However, the version of wxWindows used by the pgadmin3 > > package comes from www.pgadmin.org. I suspect I'm going to have to wait a > > bit before the whole thing will be ready for the Alpha. > > Bob, can you check wxWindows CVS head on that Alpha machine? We will > rarely update our wx package, so to get Alpha fixed the only way is to > use an official wx version and have one of their developers work on > that. It might even be a bug in gtk2, which would need an escalation to > the gtk team... > > Regards, > Andreas Andreas, Sorry it has taken me so long to respond, but we lost electrical power because of hurricane Isabel. I downloaded the file http://wxwindows.sourceforge.net/snapshots/wx-cvs-20030919.tar.bz2 and built it with the standard ./configure; make; make install approach. I used the --enable-unicode and --disable-shared flags for the configure script. I then made the samples target and ran the internat executable. It pops up a window titled "Language" with the following choices: System default, French, German, Russian, Bulgarian, English, and English (U.S.). It doesn't offer the Japanese or Georgian choices you mentioned, but it seems to run OK. I don't know what to do next. Regards, .... Bob
Bob Kimble wrote: >On Thursday 18 September 2003 12:17, Andreas Pflug wrote: > > >>Bob Kimble wrote: >> >> >>>On Thursday 18 September 2003 11:30, Andreas Pflug wrote: >>> >>> >>>>Hi Bob, >>>>It would have been quite a surprise if that sample didn't crash too. >>>>Please post that issue at wxWindow's bug system. >>>> >>>>Regards, >>>>Andreas >>>> >>>> >>>Sounds good. However, the version of wxWindows used by the pgadmin3 >>>package comes from www.pgadmin.org. I suspect I'm going to have to wait a >>>bit before the whole thing will be ready for the Alpha. >>> >>> >>Bob, can you check wxWindows CVS head on that Alpha machine? We will >>rarely update our wx package, so to get Alpha fixed the only way is to >>use an official wx version and have one of their developers work on >>that. It might even be a bug in gtk2, which would need an escalation to >>the gtk team... >> >>Regards, >>Andreas >> >> > >Andreas, > >Sorry it has taken me so long to respond, but we lost electrical power because >of hurricane Isabel. I downloaded the file >http://wxwindows.sourceforge.net/snapshots/wx-cvs-20030919.tar.bz2 and built >it with the standard ./configure; make; make install approach. I used the >--enable-unicode and --disable-shared flags for the configure script. I then >made the samples target and ran the internat executable. It pops up a window >titled "Language" with the following choices: System default, French, German, >Russian, Bulgarian, English, and English (U.S.). It doesn't offer the >Japanese or Georgian choices you mentioned, but it seems to run OK. I don't >know what to do next. > This is what it is supposed to do, so the problem seems to be fixed in the meanwhile. After the interim release of wx is out, I'll check that version (and apply our patches...), you then hopefully can run pgAdmin3 on Alpha too. Regards, Andreas