Re: Case mapping of sharp s

["Stephen J. Turnbull" <stephen@xemacs.org>]

Eli Zaretskii writes:

 > Not that it's important, but since we are talking principles here: the
 > bidi algorithm has nothing to do with Unicode, let alone UTF-8.  It is
 > not even the best algorithm to handle bidirectional text (rumor has it
 > that Microsoft lobbied the consortium into the algorithm they
 > developed for their word processors).

That would not be surprising, and it might even be the best result.
The "POSIX_ME_HARDER" argument notwithstanding, a mediocre standard is
often better than a proliferation of excellent nonstandards.

 > Better algorithms with saner results were available years before
 > UAX#9 was published, and they worked with codepoints from DOS the
 > codepage 862 just dandy.  You (Stephen) may recall that I presented
 > one of them in Tsukuba 9 years ago.

Sure, and it's still not usable in any popular editor AFAIK, while
millions of people are using the Unicode standard algorithm if it is
in fact the one the Microsoft uses in its wordprocessors.  Bash and
other nonconforming derivatives of the POSIX sh, on the other hand,
are in universal use.