Re: Build problem

[Ethan M. <merritt@u.washington.edu>]

On Monday 10 July 2006 12:43 am, Timoth=C3=A9e Lecomte wrote:
> Ethan A Merritt wrote:
> > Furthermore, to the extent that these various character sets are
> > all implemented via UTF-8,  why would one need separate modules
> > anyhow?
>
> Then, Pango and text rendering in general is not only about drawing
> individual characters. It is about glyphs, layout, typography and I
> should maybe add calligraphy. And this goes very language-specific.
> Arab for example needs the glyphs to be joint together, and the same
> character at the end or in the middle of a word would be drawn as a
> different glyph.

That make sense.  I am not familiar with the orthography of all the
modules in that list, but it does seem likely that some
languages would require extra rules for placement and connection
of adjacent letters.  Of course, even English othography does that to=20
some extent via font kerning.

> Maybe Chinese and Japanese don't need those special
> modules, because the typography is closer to occidental languages
> (one character in utf8=3D one glyph in the output, always the same ?).

Some fonts vary by a factor of 2 in character width, but I doubt that
simple test would require a specialized module.  On the other hand,
x11 gets it wrong without external help, so maybe it's trickier than
it would seem.


=2D-=20
Ethan A Merritt
Biomolecular Structure Center
University of Washington, Seattle WA