Re: [PATCH 0/6] cover glyph support
Behdad Esfahbod <[email protected]>
| Newsgroups | gmane.comp.lib.cairo |
|---|---|
| Message-ID | <CAF63+7UH+k25VOEY5i-b=QSoEZXT6QOXejduZv+EvXMoEjCH2A@mail.gmail.com> |
On Sat, Apr 9, 2016 at 7:39 AM, Matthias Clasen <[email protected]> wrote: > On Sat, Apr 9, 2016 at 3:54 AM, Chris Wilson <[email protected]> > wrote: > > > > It's not mergeable since you break cairo_show_glyphs(), which now > > randomly operates differently based on whether the system loaded a > "color" > > glyph - your implementation completely discards the source and changes > > the definition of the glyph operator. > > > > To put this another way, there are several places in cairo (and > pixman) which make the assumption that "fonts are collections of > masks". > Fonts with color glyphs invalidate this assumption, which is why these > patches are a bit ugly. > > Do you have advice on how to handle this nicely in the cairo architecture ? > It was part of cairo's original imaging model that show_glyphs is one of the various methods to specify an area within which a Porter-Duff operator happens. It. it was defined as a mask. Color fonts change all of that; I remember, I think it was Owen, who was opposed to making such changes to show_glyphs when I first brought up color fonts years ago. But the world has moved in this direction, and I don't see any other option. Asking the cairo client to separate color glyphs and draw them using a different operator is not an option IMO. That said, perhaps we should add a flag to scaled_font_options as well. -- behdad http://behdad.org/ -- cairo mailing list [email protected] https://lists.cairographics.org/mailman/listinfo/cairo