[cairo] Any objections to dropping _cairo_atsui_font_old_show_glyphs?

Baz brian.ewins at gmail.com
Fri Mar 16 15:29:20 PDT 2007


Great. Consider it gone!

On 16/03/07, Vladimir Vukicevic <vladimir at pobox.com> wrote:
> Behdad Esfahbod wrote:
> > On Wed, 2007-03-14 at 12:28 -0400, Baz wrote:
> >> Just looking for comment on this. _cairo_atsui_font_old_show_glyphs
> >> was used to display glyphs under the old quartz code, for all
> >> surfaces. More recently its become unnecessary for any surface and in
> >> fact only gets used in these (rare) circumstances:
> >>
> >> - atsui text is being rendered to a quartz surface
> >> - and we cannot render the text in cairo-quartz-surface (this happens
> >> because, eg we cannot convert a pattern to a CGShadingRef);
> >> - and we can extract a destination image surface from the quartz
> >> surface. The code looks like the original quartz surface could be
> >> used, but this is actually unreachable.
> >> - and we can create a CGBitmapContext matching the destination image
> >> surface. This should always succeed.
> >>
> >> If any of that fails it ends up using glyph surfaces, which should
> >> work anyway - they are used for all non-quartz surfaces.
> >>
> >> Does anyone see this code[1] getting used in the real world? It isn't
> >> being hit by any of the tests any more, but of course its still just
> >> as buggy as it was with the old quartz :). If no-one is actually using
> >> it, I think we'd be better off without.
> >>
> >> The cut would just be 150 lines or so. Some of this comes from
> >> inlining what little of of cairo-atsui's
> >> CGBitmapContextCreateWithCairoImageSurface is still required.
> >>
> >> Cheers,
> >> Baz
> >
> > Go ahead from me!
> >
> > Vlad, comments?
>
> Killing it sounds great to me!
>
>      - Vlad
>


More information about the cairo mailing list