Martin Sevior wrote: > This is a quick fix for the two non-Latin fonts current distributed with > Abi. As abi gains more non-latin fonts we should implement a more general > way of fixing this. Maybe checking to see if a given font is in the list of > Latin fonts. If it is issue the LAT. If it isn't don't issue the LAT. Yes, the general problem is that we need to match the encoding to the font, and do the right thing for each file. I'm not sure of the standard way to detect encodings (should we use XLFD strings and ignore the font files?). > It also issues the "stroke" command to fix the variable width underline > printing. I had done a little research into this today. I haven't seen your patch yet (will decompress after this letter). It seems sufficient to include a "stroke" after the "lineto", but my knowledge of the mapping from "a GC's function to draw lines" to "PostScript fills these pixels" doesn't include the differences between stroked paths (for drawing) and a simple lineto. -- Shaw Terwilliger
