On 02/08/2002 03:01:31 AM John Hudson wrote:
>Chris, how do you define a 'properly set' Unicode range in the OS/2
table?
>
>Correct codepage support is self-evident: a font should indicate codepage
>support only if it's cmap table includes *all* the characters in that
codepage.
Well, there are some gray areas. There are fonts out there that have the
bit for cp1252 set but that don't have the euro or the upper/lower
z-caron.
And, I will confess, there are fonts out there that really stretch their
claims to supporting cp X. For example, when we were completing our Yi
font a couple of years ago, we wanted it to work in Word 97 and Word 2000.
There was a problem in that Word 2000 had a bunch of font-linking things
going on to try to keep the user from seeing boxes, but the algorithms
were completely unaware of Yi. I ended up having to set codepage bits for
Japanese and (I think) Central European (some Latin codepage other than
cp1252) in order to make Word 2000 actually use the font -- if I didn't,
then Word would quietly substitute Times New Roman or a Far East font for
characters that the font really did support, including about half the Yi
range. The claims of supporting those two codepages was very tenuous:
there were many of the cp1250 characters not supported by the font, and I
think there was exactly one character from cp932 that we actually
supported -- 30FB. I'm sure we're not the only ones who have ever
stretched things like this.
- Peter
---------------------------------------------------------------------------
Peter Constable
Non-Roman Script Initiative, SIL International
7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA
Tel: +1 972 708 7485
E-mail: <peter_constable@sil.org>
This archive was generated by hypermail 2.1.2 : Fri Feb 08 2002 - 11:14:53 EST