>From: Owen Taylor <[EMAIL PROTECTED]> > > How are you encoding the font? > > I think the only long-term sensible encoding for either Postscript or > TrueType outlines in an Indic script is OpenType with the correct > GSUB/GPOS tables, though clearly figuring out how to create such > a font is a large amount of work. (The Indic OpenType spec is > available from Microsoft's web site.)
I am open to suggestion. From my research, it seems that the problem is in three parts: the encoding rules themselves, where to store them, and a lack of code points for the ligatures. For the encoding rules, if the PANGO_LIGATURE_HACK work has produced working algorithms, then I am sure that they can be transcribed into virtually anything else that is required. There also seems to be a body of work done by the ITRANS project along these lines. There is also some work apparently being done on a "library to write Bengali". For where to store the encoding rules, it seems to me that they are best carried along with the fonts: this gives the flexibility to have fonts in different stages of completeness. I don't know if X can handle OpenType fonts...if not, then I something like the PANGO_LIGATURE_HACK might be the way forward for now. For the ligature glyphs, there seems to be a strand of thought suggesting the use of the the private space within Unicode for additional ligature glyphs. (BTW: for the moment, I am concentrating on the glyphs specified by Unicode). > I don't know if Qt3 will handle OpenType / Indic fonts. We have > most of the infrastructure to handle them for Pango and GTK+-2.0, > though the Indic-specific pieces stiill need to be implemented. There have been some noises from Norway :-), but I have not heard of any concrete plans at this point. > Without a standard encoding that can be used by available software, I > don't know if a collection of glyphs is that interesting and addition > to XFree86, though it's clearly an essential step to getting > appropriate fonts in the end. I couldn't agree more - the glyphs are something that I can make a difference with - but we need to get an agreed solution to the encoding problem. BTW, although I would share your implicit concern not to paint ourselves into too much of a corner, I would rather see something that is 80% done, and which can be extended to say 98%, than wait much longer for perfection. I suspect that bringing together the available work to date will get us there. Thanks, Shaheed _______________________________________________ Fonts mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/fonts
