From: Peter Constable (firstname.lastname@example.org)
Date: Tue May 25 2004 - 16:18:09 CDT
> From: email@example.com [mailto:firstname.lastname@example.org]
> Of Dean Snyder
> Archaic Greek exhibits variable glyph stance, that is, glyphs can be
> flipped horizontally or even vertically, usually dependent upon the
> direction of the writing stream.
> How should variable glyph stance for the same characters in the same
> script be dealt with in Unicode and in a Unicode proposal?
If you're talking about what happens in boustrophedon text,
then I'd treat it as a presentation issue, not an encoding issue. IMO,
it would be a serious problem if you have to encode an alpha using a
distinct character just because it happened to come (with a given text
size and page metrics) on the RTL run of boustrophedon layout rather
than a LTR run. At the *very most*, you might propose control characters
that can be used to distinguish whether characters in a given run of
text should be rotated or mirrored if part of a RTL line, but even there
I would be inclined to leave that to higher-level processing and
If you're talking about variations among archaic documents in how
particular letters are written, apart from line direction issues, e.g.
then you might have a case for proposing variation-selector sequences.
Globalization Infrastructure and Font Technologies
Microsoft Windows Division
This archive was generated by hypermail 2.1.5 : Tue May 25 2004 - 16:19:01 CDT