On Sat, 30 Jun 2001 06:26:42 -0500 Michael Kaply <[EMAIL PROTECTED]>
wrote:
>
> On Windows and every other platform except OS/2, it is possible to query
> any font to see if it supports a particular glyph. So all Windows does is
> ask the font "do you have this glyph?" If it does not, it finds one that
> does. We can't do that on OS/2.
Can one on OS/2 find out whether the Times New Roman MT30 font has
been specified for Mozilla ? If yes, may I suggest testing for it,
and performing "special" rendering for just that one font.
I believe Mozilla/2 already has some "data-dependent" tests (such
as for a '.pdf' filetype, even when no "helper" entry has been
specified for that type). The test for MT30 would be one more
such bell/whistle.
The advantage would be that users who CHOSE to specify MT30 to
Mozilla/2 as the font for 'Western' (or whatever) *would* get
to see the large number of glyphs in that font, even when the
page was 'UTF-8' being treated as 'Western'. (If the user did
*not* specify MT30 for 'Western', then glyphs would be rendered
through the existing Mozilla display mechanisms.)
>
> So when you view a Russian page, we convert the Unicode
> that Mozilla sends us into Win1251 and display it. We have tried displaying
> everything in UTF-8, but it does not work. The reason is because most OS/2
> fonts do not support UTF-8, so you just get junk. We would have to display
> every page in Times New Roman MT 30.
My suggestion: Wherever the user has specified MT30 in the 'Fonts for'
entry, __do__ display in unicode. Wherever the user has __not__
specified MT30, display it as you do today.
[Then the USER decides if he wants to see the information in MT30 or not.
I myself *have* specified the MT30 font to be used for the various
character sets I use (including 'Western').]
mikus
