[Fenfire-dev] gl loom

[Matti Katila]

Hi, 

I tested shortly loom with GL and vobscene creation time was from 200 
to 500 milliseconds. I don't think that we should try to put our 
energy to optimize anything from c++'s side. Yeah, it's not fast enough 
but that might take too much time even try to figure out the 
bottlenecks.

I also think that we have not got enough machines to make such a test 
where every machine would perform well. Like I said, 2.6 kernel has 10 
times faster context-switching and it might help a bit. We may try it 
next.


   -Matti