Hi, Excuse me, could you post some pictures/screenshots comparing the expected older result & unexpected newer result?
Regards, mpsuzuki On Fri, 17 Sep 2010 09:19:18 +0400 Ineiev <[email protected]> wrote: >Hello, > >Recently I noticed a change in default FreeType rendering >of one of the fonts I use; the change was not desirable >for this font, so I tried to reproduce the older behaviour. > >I tested different configuration options from ftoption.h, >each time opening the font with FT_PARAM_TAG_UNPATENTED_HINTING, >as documented in ftoption.h, and without it. > >I distiguish the interpreters by the form if the >outline: the 'unpatented' one produces a regular >rectangle (which is the desirable result), >the 'patented' one emits a trapezium. > >The test program source and it's output is attached. >The command line was './dec lohit_hi.ttf 109', the font file being >http://savannah.nongnu.org/bugs/download.php?file_id=21450 > >It appears that: > >When configured for FT_TRUETYPE_ENGINE_TYPE_NONE, it uses >the 'unpatented' bytecode interpreter in both cases; > >When configured for FT_TRUETYPE_ENGINE_TYPE_UNPATENTED, it uses >the 'unpatented' bytecode interpreter when >ft2-truetype_tables.html#FT_PARAM_TAG_UNPATENTED_HINTING >documents the 'patented' one and vice versa; > >When configured for FT_TRUETYPE_ENGINE_TYPE_PATENTED (current >default), it uses the 'patented' bytecode interpreter in both cases. > >It means that either I do something wrong again (after all, >ftview shows rectangle in all configurations) or newer FreeType >versions in default configuration are uncapable to render >'certain Asian fonts' as they used to; >in the latter case, I'm really not sure what is intended, >but probably there should be more documentation details about it. > _______________________________________________ Freetype mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/freetype
