discuss-gnustep
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Q] What happened with subpixel rendering in art backend?


From: Johannes Fortmann
Subject: Re: [Q] What happened with subpixel rendering in art backend?
Date: Tue, 11 Oct 2005 17:29:58 +0200


Am 08.10.2005 um 15:11 schrieb Christopher Armstrong:


Having read the WWN (wine weekly news) lately, I think I can answer your
question. Subpixel rendering in freetype2 uses algorithms under patent
(ironically, their held by Apple) and freetype2 is usually compiled
without them, to avoid infringement. If you live in a country that does
not recognise software patents, you may be able to recompile freetype2
for subpixel support.

Apparently the wine project is having similar issues, and is looking for
a solution as well.

This is coupled from what I remember reading on WWN; you may want to
check out freetype2's mailing list for more details. I'd say this is
what happened to subpixel rendering.

Cheers

I don't think that is the issue here. freetype2 basically has two modes:
-with TT_CONFIG_OPTION_BYTECODE_INTERPRETER defined in ftoption.h,
the Apple-patented bytecode interpreter is used. This means type hinting is done
 according to the hinting information in the font.
- without that option set, there are no hints. That doesn't keep freetype from anti-aliasing, though: it just has to guess what the hinting would be (this is done by the so-called "auto-hinter"). In some cases, this option can even hint much better; there are some fonts with corrupt hinting information floating around, which look horrible
with the interpreter option.
Anyway, the worst that can happen if the bytecode hinter is disabled is that the glyphs look a bit worse at small sizes (<10 pt.). The _availability_ of antialiasing
should not be affected, only the quality.

BTW, the really ironic part is not that Apple holds the patents (after all, it would be them or MS), but the fact that Mac OS X ships freetype2 with the interpreter off.

HTH,
J. Fortmann




reply via email to

[Prev in Thread] Current Thread [Next in Thread]