Solaris MH direct compiling

[Sergio]

Hi all,

Can sameone help me about direct compiling MH (svn) : on Solars x86, i can build easily MH, but at the execution it crashes immediately at each time.

-32 or 64 bits ;
- with mesaOpenGL or NvidiaOpenGL ;
- installing pyglew or not ;
- in all cases, with tthe latest SDL and the latest GLEW.

I'm currently using MH 0.9*, never problems like this on Solaris x86 or Sparc.

Cheers,

Sergio

[mflerackers]

do you have a gdb stack trace or python output?

[Sergio]

Hi mflerackers,


Thank's !

Hire my exact gdb backtrace output :

bash-3.00# /opt/csw/bin/gdb makehuman core
GNU gdb 6.6
Copyright (C) 2006 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 "i386-pc-solaris2.8"...
(no debugging symbols found)
Core was generated by `./makehuman'.
Program terminated with signal 11, Segmentation fault.
#0 0xfefeb965 in main ()
(gdb) bt
#0 0xfefeb965 in main ()
(gdb) quit
bash-3.00#


Other hand, what kind of python output can i provide ?


Cheers,

Sergio

[Sergio]

OK, seems for the moment solved : the problem came from my distribution libpython2.6, maybe because it's only as static library proposed. Strange as a reason, assuming that the build itself is possible with the -shared ld option... In all cases, building SVN MH with my old Python-2.5, it seems normally working. But maybe i'll loose some MH features ? Can anyone confirm or infirm that ?

Cheers,

Sergio

[mflerackers]

How did you build? If you used scons, I would think it gives an error. MakeHuman doesn't use any 2.6 specific constructs yet, though this may change in the future.

[Sergio]

Hi mflerackers,

Without Scons, just using the Makefile with my own settings. The build was proper, now my only one problem is to known if Python-2.5 can be sufficient for all the actual MH features, or if i must build especially a Python-2.6 or 2.7.

Cheers,

Sergio

[jcapco]

why do you give Sergio a headache.. of course there won't be any python 2.6 specific things. Go on continue with 2.5 .. I recommend 2.5 until ironpython, scipy and numpy becomes python 3.0 compatible.

[joepal]

Hi mflerackers,

Without Scons, just using the Makefile with my own settings. The build was proper, now my only one problem is to known if Python-2.5 can be sufficient for all the actual MH features, or if i must build especially a Python-2.6 or 2.7.

Cheers,

Sergio

I think osx is still built with 2.5? Anyway, I never had any problems using 2.5 for the linux builds until I re-standardized on ubuntu 10.04 which had dropped support for 2.5.

[Sergio]

Hi joepal,

Thank's greatly ! And you are right, i omitted to take a look in the Makefile.osx...

Cheers,

Sergio

[Sergio]

Hi jcapco,

Thank's also ! Okay, i'll continue with my working 2.5...

Cheers,

Sergio