On Thu, Mar 10, 2016 at 9:53 PM, Keith Packard <keithp@xxxxxxxxxx> wrote:
Behdad Esfahbod <behdad@xxxxxxxxxx> writes:
> I've started to convince myself that fontconfig should NOT by default cache
> the charset info of all fonts, and do so only lazily. This will solve most
> of the "caching is slow" issues... Needless to say, that's better done in
> a fontconfig3...
How would it perform font matching based on language then? Ask for a
Japanese 'sans-serif' font and you'll get dejavusans...
For system fonts, the cache will be as it is right now. For anything else, user processes just compute the coverage the first time it's actually needed. That means the lang/charset scoring needs to be lazy. The point being that a painful majority of times, the match is clear already based on family name matching, and the fact that the matched family supports the language requested.
We still be doing all the things we do right now. Just postpone computing the charset to the first time it's actually needed, and add it to the cache then.
--
-keith
--
behdad
http://behdad.org/
http://behdad.org/
_______________________________________________ Fontconfig mailing list Fontconfig@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/fontconfig
