[Top][All Lists]

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

Re: [ft-devel] otv in Jamfile

From: Werner LEMBERG
Subject: Re: [ft-devel] otv in Jamfile
Date: Thu, 02 Jun 2005 09:01:02 +0200 (CEST)

> The otvalid module doesn't compile at all with Visual C++, due to
> its liberal use of subtle but legal preprocessor tricks (blame it on
> VC++, not FreeType).

Have you tried the sed script which I posted on this list?  Using it
to expand the affected files should allow a clean MSVC build.

> However, I believe it's not a good idea to put such a module within
> the font library. As Owen Taylor pointed it out, that's because if
> you implement a library to process the OpenType tables, and if some
> of these are updated or added by the specification, you don't want
> _your_ code to depend on an updated version of the FreeType
> validator to work correctly.

But the same argument holds if any of the font formats FreeType
supports is updated!  The otvalid module handles OpenType 1.4 --
nothing stops the user to take the otvalid code, put it into the
application and update it there if necessary.

It's OK with me to put otvalid into another library, but unfortunately
we don't have the proper higher-level library yet which would be the
right place.

Just to be sure I'll add a note to the documentation that this module
might be moved to another library in the future.

> By the way, could we just stop adding new features.  We're supposed
> to be in the freezing stage, and I doubt that the new emboldening
> code is bug free at the moment.  Can't it be held back for a bit
> later ?

Basically, you are right, but the code Chia-I has contributed is on
the border line IMHO: While extending the emboldening code he has
fixed some serious bugs in it.  Additionally, his changes are natural
and give better orthogonality.

BTW, today or tomorrow I'll handle the `const' problem, then we can do
a final pretest with a complete code freeze until 2.1.10 is out.


reply via email to

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