From: Doug Ewell (email@example.com)
Date: Tue May 04 2004 - 10:19:22 CDT
Philippe Verdy <verdy underscore p at wanadoo dot fr> wrote:
>> A problem, however, is that many such forms are found in unstable
>> orthographies, and are difficult to document adequately for inclusion
>> in proposals.
> This last argument should not be a limitation to encode them. After
> all they are used for living languages in danger of extinction, and
> even if documents using them are rare, encoding them would help
> preserving these languages and helping the development of their
This is expressly NOT a goal of Unicode and ISO/IEC 10646: to encode
newly invented, possibly ephemeral, letters on the basis that doing so
might encourage literacy and save a language from extinction.
As someone once said -- I don't know who, but it sounds like John
Cowan -- we already have several hundred Latin letters in Unicode; it
shouldn't be difficult to pick one of those when developing a new
orthography, instead of inventing yet another way to write [tʃ].
The danger of encoding novel characters on speculation that they might
be useful is that if they *don't* turn out to be useful, or if a revised
version of the orthography replaces them with something else, Unicode
and 10646 are stuck with unwanted characters, which cannot be removed
for stability reasons.
The Euro sign is a classic counterexample where strong promises of
stability and usefulness (which have been amply borne out) outweighed
the "newly invented" nature.
See the Principles and Procedures document for more information.
> Without them, the instability of orthographies will always be a
> problem favored by absence of standard to represent them adequately in
> any encoding or charset, so that even book publishers and authors will
> need to use their own approximations or unstable private conventions
> to represent them.
This is a problem; in an increasingly Unicode world, it is more
difficult than ever to print and interchange one's characters if they
are *not* in Unicode. But the burden should still be on the proponents
of such a character to prove that it is in actual, stable use, and that
the need to print and interchange is real. Otherwise, it's PUA time.
This archive was generated by hypermail 2.1.5 : Fri May 07 2004 - 18:45:25 CDT