kicad-developers team mailing list archive

[Dick Hollenbeck <dick@xxxxxxxxxxx>]

On 04/23/2013 03:00 PM, "Torsten Hüter" wrote:
> Hi Maciej,
>  
> thanks for your efforts, nice work, I'll try this again for various boards. Indeed I've
> spotted the missing callback function as well.
> It's the combine callback of the tessellator, but this one is only called if there is an
> self-overlapping polygon.
>  
> So it should be clarified why there is such a polygon existing; if the zone filling
> algorithm works correctly it shouldn't exist.
>  
> I'd disagree with Lorenzo, as the originator of the GAL I'd suggest that you stick with
> the first plan, that means using display list first.
> As you have (perhaps) already seen, you can use the group methods for that:
> BeginGroup(..)
> EndGroup(..)
> DrawGroup(..)
> DeleteGroup(..)
>  
> VBOs are a kind of optimization and I'd implement it, after everything else works fine.
> The reason is, that it requires much more code for the same functionality. For instance
> you need to keep track of the various objects, like triangles, triangle strips, quads and
> you need to organize the memory transfer etc. At this point I'd implement also the major
> drawing routines as shaders using the interface that I've provided. But I'd do that later.
>  
>  
> The question  for me in general is the minimum required OpenGL version, this is important
> for shader handling and other features.
>  
> --
>  
> Can you use the current location of the GAL for your bug fixes, this would be really nice,
> so that I and Dick for instance can commit as well:
>  
> lp:~kicad-testing-committers/kicad/kicad-gal


Torsten,

Perhaps it would be easier for you to be added to the cern-kicad group.  I suspect the GAL
stuff was brought into the the tree because of convenience.  (This was what you were
working for right?)

As for myself, I do not foresee me needing to commit anytime soon to GAL.  I have a number
of patches to look over and work to do on python for windows, fp lib table dialog perhaps.

Dick