Can't compile simple.cpp with event handling.

[jorrit]

Very strange. I don't have an explanation and I haven't heard from  other people about this. So at the moment I can't give you any hints on how to solve this I'm afraid

Greetings,

[Winter Knight]

With your help, Jorrit, I solved the original problem, that of not being able to get simple.cpp with event handling to compile. The problem was, that when coming up with the link line, I only added `cs-config --libs` to the link line I already had. I didn't think that adding extra library lines to the link line would hurt, but apparently it does. I think it is because, like you said, the linker is one pass. Note that I only got pseudo-stable crystal to work. SVN crystal still causes simple programs to segfault. (SVN as of this morning PST). I'll be checking out SVN again in a week or two.

Speaking of SVN, did the gdb info I sent help?

EDIT: When I wrote this, I didn't see your previous message. 2 pages...

[Winter Knight]

Checked out svn again a few hours ago. And I compiled using gcc-3.4 (latest debian sarge version, 3.4.3), rather than gcc-3.3. I noticed in the manual that it suggests 3.4 or above, but it does not say that 3.3 is unsupported.

3.4 makes a world of differnce. Yeah, there are still errors, glitches, and bumps. But, if you compile svn with 3.4, walktest will run. It's a castle. In spite of the fact that the castle appears to be simpler, it is much, much slower than the map that walktest uses by default in pseudo stable. It might be the errors. I have them here:

wk@ALPHA:/mnt/hdb2/svn/crystal$ ./walktest -relight -fs
DEBUG: Sound System Software Renderer Initializing...
WARNING! Object 'Cube.582' is not closed!
WARNING! Object 'Cylinder.232' is not closed!
WARNING! Object 'Cylinder.230' is not closed!
WARNING! Object 'Cylinder.012' is not closed!
WARNING! Object 'Cube.580' is not closed!
WARNING! Object 'Cube.578' is not closed!
...
Total level load time: 5.682 seconds
Cleaning up...
X Error of failed request:  BadWindow (invalid Window parameter)
  Major opcode of failed request:  4 (X_DestroyWindow)
  Resource id in failed request:  0x1200003
  Serial number of failed request:  133
  Current serial number in output stream:  135
wk@ALPHA:/mnt/hdb2/svn/crystal$

I'm still new to crystal space, so I don't know if those are serious or not. Anyway, that is latest svn walktest running from installation directory, where program is "jammed" in the source tree. Actually installing crystal space will always break it, at least with every test I've done.

Jorrit, I've asked you before what version of gcc you use, and you didn't respond. I can respect your reasons, whatever they might be. Are there any developers willing to share what environment they develop and test crystal space on/for? That would be really helpful for those of us using different environments. Thank you.

[jorrit]

Jorrit, I've asked you before what version of gcc you use, and you didn't respond. I can respect your reasons, whatever they might be. Are there any developers willing to share what environment they develop and test crystal space on/for? That would be really helpful for those of us using different environments. Thank you.

I didn't respond because I didn't notice the question. Sorry. But I do use gcc 3.4

Greetings,