On 08/13/2010 12:30 PM, Benjamin Root wrote:
>
[...]
>
> Notice that both stroke and fill are checking for alpha != 0.0.
>
>
> Yeah, well, try out my attached script. Then view the two files.
> Something is wrong...
I think I have that fixed now, along with a temporary fix for the
original "none" problem. The better fix for the latter, involving
to_rgb_array, can be done in the trunk.
The handling of alpha, and colors in general, is still confusing,
providing abundant hiding places for bugs and opportunities for
inefficiency. I hope that over time we can clarify and codify it.
For example, in some cases in the backends, the alpha value of an rgba
does serve as a flag; in other cases, it is simply ignored, and the flag
value is python None.
As far as I can see, at least for a Patch, one can explicitly set
different alpha for the outline and for the face, but the backends can't
render them that way. I don't see any reason why we can't fix that, but
it would take some backend work. It might be fairly simple to factor it
out into a pre_draw_path() that would call the backend separately for
the edge and for the face if they have different alpha values, but could
use the present single call otherwise. I suspect this could simplify
the backends, relieving them of some of the checking they are doing now.
Eric
>
> Ben Root
--
This SF.net email is sponsored by
Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev
___
Matplotlib-devel mailing list
Matplotlib-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
