Re: BALLView - a molecular viewer and modeling tool

[Andreas Moll <amoll@bioinf.uni-sb.de>]

Teemu Ikonen schrieb:

> The code above causes a build failure in 32-bit powerpc, the uname -m
> output in this case is 'ppc'. The attached patch fixed this for me.
Hi,

thank you very much for the patch, it will be included in the next 
package version.

> I also built ballview on a 6-month old (i.e. fast) pc, but it seems
> really slow, the screen update is something like 0.3 FPS. I have an
> intel graphics card with DRI enabled and other opengl apps work ok.
> Any idea what might be wrong?
Most intel graphics card are not really renowned for their performance.
Or maybe you have a driver problem?
Just for comparision, how many FPS can you get in glxgears?

 http://www.elsniwiki.de/index.php/Main/Glxgears

You may also try to run the opengl_benchmark.py under 
/usr/share/BALL/PYTHON.
With a 5 year old Geforce 4TI I get a result of 2.28553815935 BALLView 
OpenGL stones and it can fluently render stick models or surfaces for
molecules like 2PTC with more than 2000 atoms.

For a test you may also try to disable DRI and see what happens...

And if nothing else should help, you can switch to a lower detail level
in BALLView. In the lowest level, it will look choppy, but should at
least give you enough FPS.
As an alternative use the line model.

I hope these tip will help you along...

Best regards

Andreas