[..] > The data provided to apt could be as coarse or as fine as needed; font > supports unicode scripts xyz, font supports unicode blocks xyz, font > supports unicode codepoint ranges xyz. > > BTW, how does the RPM world propose to implement this? AFAICT it's pango/freetype/fontconfig underneath which isn't distro-specific. fontconfig has "orthographies": i.e. matching of codepoints to languages: http://cgit.freedesktop.org/fontconfig/tree/fc-lang Nicolas Mailhot - Fedora font packager extraordinaire - would probably know if they intend to rely on extra info in the packaging metadata or simply grab the font coverage info from the fonts automatically and put them somewhere else. Until we redo many of our open font packages maybe we could have a separate database with this info? Somehow put fc-lang output in our review script, maybe? Not quite sure how the fallback mechanism is designed to work... > Also, what data does packagekit provide to the distro backend? Unicode > blocks? Lists of Unicode code points? > >> Cheers, > > Thanks for working on packagekit! Indeed a *huge thank you* for your work on this. This is incredibly useful :-) Cheers, -- Nicolas Spalinger, NRSI volunteer http://planet.open-fonts.org
Attachment:
signature.asc
Description: OpenPGP digital signature