Hi Lars,
The code in oconv checks modifier and string lengths, though there may be problems in xform with long names. That shouldn't crash oconv, however. It's difficult to guess without a traceback.
-Greg
···
From: "Lars O. Grobe" <grobe@gmx.net>
Date: July 25, 2009 9:15:45 AM PDT
Hi Greg!
Your backtrace is missing symbolic information telling what call was
last made in Radiance. It seems to have crashed in a sprintf()
statement, which means it could have bombed while trying to print an
error message. You should recompile with the -g option and crash it again.
Well, this is the ONE time that I had not compiled myself, so I will see
whether I can reproduce this.
Just as a guess, you wouldn't happen to have any strings longer than 127
characters in your material descriptions, would you?
No, but possibly in file names included by xform, in mesh objects and
such. Errrrr does your question mean that there is no overflow
protection for modifier names that exceed the maximum string length?
Currently downloading the latest snapshot for a -g build 
Cheers, Lars.
Lars O. Grobe wrote:
Currently downloading the latest snapshot for a -g build
If I remember right there is some way to retrieve the external debug symbols in
Ubuntu... But I could add a -dbg package in any case.
If you want to give the latest Debian package a try and compile it with debug
symbols, do the following:
git clone git://git.recluse.de/debian/pkg-radiance.git
cd pkg-radiance
export DEB_BUILD_OPTIONS=nostrip,noopt
git-buildpackage
(if you don't have git-buildpackage, use dpkg-buildpackage -i.git -I.git)
That should give you some .deb files with included debug symbols.
Then try to get a core dump, or run the executable in gdb.
ulimit -c unlimited
will help:)
In gdb please use
bt full
to get a full backtrace.
Then file it as bug in the Debian bts or send it to this list 
Cheers,
Bernd
···
--
Bernd Zeimetz Debian GNU/Linux Developer
GPG Fingerprints: 06C8 C9A2 EAAD E37E 5B2C BE93 067A AD04 C93B FF79
ECA1 E3F2 8E11 2432 D485 DD95 EB36 171A 6FF9 435F
Currently downloading the latest snapshot for a -g build
Hi.
for your info, the binaries I created run fine. Obviously something is wrong with the binaries Ubuntu provides, at least for amd64. So this topic is not for the radiance-dev list but for Ubuntu.
Cheers, Lars.
Hi!
Lars O. Grobe wrote:
for your info, the binaries I created run fine. Obviously something is
wrong with the binaries Ubuntu provides, at least for amd64. So this
topic is not for the radiance-dev list but for Ubuntu.
As nobody will take care of bug reports in Ubuntu (as nobody knows about the
package and there is no real maintainer) - you'll have to stick with me 
If the Debian package works and the Ubuntu one not - there is not much I can do
except asking for a rebuild.
Please build radiance from my git as described before and see if the bug is in
there.
Cheers,
Bernd
···
--
Bernd Zeimetz Debian GNU/Linux Developer
GPG Fingerprints: 06C8 C9A2 EAAD E37E 5B2C BE93 067A AD04 C93B FF79
ECA1 E3F2 8E11 2432 D485 DD95 EB36 171A 6FF9 435F
Hi,
for your info, the binaries I created run fine. Obviously something is
wrong with the binaries Ubuntu provides, at least for amd64. So this
topic is not for the radiance-dev list but for Ubuntu.
could you figure anything out? Ubuntu is starting to freeze again, if there is
something to fix, it should be fixed now.
The CVS head from 20090601 is in Ubuntu now, which is pretty old already
compared with what I have in Debian.
Greg: I've seen that you were doing various minor changes during the last weeks
- when is a good day to get your CVS Head and push it to Ubuntu? I think you've
tried to fix Lars' bug !?
Cheers,
Bernd
···
--
Bernd Zeimetz Debian GNU/Linux Developer
GPG Fingerprints: 06C8 C9A2 EAAD E37E 5B2C BE93 067A AD04 C93B FF79
ECA1 E3F2 8E11 2432 D485 DD95 EB36 171A 6FF9 435F
