Source: Mark Davis
Date: 2006-10-16
Subject: Yiddish Digraphs

Cary raised the issue of confusable Hebrew characters.
The characters in FB00..FB4F all have compatibility decompositions.
(Such characters are included for compatibility, but should be avoided
where not needed.) They also do not survive a passage through NFKC,
and are thus already not included in the output of IDN -- so they
don't currently represent any problem.

The Yiddish digraphs at 05F0..05F2 are another story; they are clearly
confusable -- and not accendentally -- they are simply a combination
of two characters. As I recall, we included them because they can
have vowels applied to them as a unit (but my memory might be
inaccurate). Unfortunately, we didn't make them compatibility
characters, so according to Cary (and I can well believe it)
we are seeing Yiddish spelled in both ways, which causes problems
for searching and other comparison.

The UTC should add these characters to the confusables, and clarify
whether they should be used in Yiddish, or their components should be.

--
Mark