Keith Packard wrote:
> To generate language coverage for a font, I need to know what Unicode
> coverage is required for each language.  I don't want the coverage
offered
> by fonts designed for the language; that's often far broader than the
> coverage needed to display text in the language.  All I want are the
> Unicode codepoints for the alphabet, abjad or logography, that way fonts
> are strictly selected based on language coverage and ignore spurious
> punctuation or foreign characters common in encodings.
> They should be as complete as possible;
> the goal is to avoid using fonts which are missing some common
codepoints,
> one such example is attempting to use an ISO Latin-1 font for Turkish;
> Latin-1 has all but two codepoints needed to display Turkish, making it
> nearly complete but also completely unsuitable.  Here's an example which
> should make the format abundantly clear:
> # Dutch (NL)
> 0040-005a
> 0060-007a
> 00c4
> 00cb
> 00cf
> 00d6
> 00dc
> 00e4
> 00eb
> 00ef
> 00f6
> 00fc
> #0132-0133      # IJ and ij ligatures

Why is 0060 necessary for Dutch (NL) but not 00a4 ?
I can just about understand 0040, since an email address *might*
be more important than the price of something, but a back quote ?
Please require the Euro sign in euroland, unless you really are
going to drop *all* punctuation including the dollar and the "at"
in email addresses.

I'm also uncomfortable about dropping requirements for numerals;
they are more like letters than punctuation.
No font is actually going to exclude western numerals, but
a font which excludes the numerals of a language/territory
will fail to convey the full meaning of many texts.

-- 
Dr. Andrew C. Aitchison         Computer Officer, DPMMS, Cambridge
[EMAIL PROTECTED]   http://www.dpmms.cam.ac.uk/~werdna

_______________________________________________
Fonts mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/fonts

Reply via email to