[cairo] State of Win32 backend?
otaylor at redhat.com
Wed May 11 16:15:45 PDT 2005
On Mon, 2005-05-09 at 20:10 -0400, Owen Taylor wrote:
> If you have commit access (if you don't, you should :-), go ahead and
> commit. Otherwise, can you provide the patch as an attachment?
I've committed your patch now, with some extra cleanups that were
needed for the const matrix changes earlier this week and a
bug fix or two.
While I was at it, I spent some time on the optimization of
fill_rectangles() ... I actually dug up the definitions of all
the operators and figured out what combinations of:
<operator> + <constant source> + <no alpha dest>
give a solid colored destination.
Seems to work well with cairo_snippets ... I haven't tried with
GTK+ yet, but hopefully we should be using FillRect() for
all the constant color backgrounds now.
(The optimization could actually be pulled up into gstate
so that fill_rectangles() is called with operator==SOURCE as
frequently as possible... there always is a big question of
where to put optimization. Though it's not completely simple,
since we don't have a query for whether a surface has alpha
or not at the moment.)
More information about the cairo