Re: [idn] utf8/legacy versioning

[Kenneth Whistler <kenw@sybase.com>]

I'm not stepping into the issue of how to define the
charsets, but to answer Soobok Lee's specific inquiry:

> Everyone knows UCS(ISo10646) and Unicode (UTC) changes and expands over time.
> Does  UCS (ISO10646) versioning  strictly follow  Unicode Versioning (UTC) ?

The answer to that question is no. Unicode versioning is specified exactly
on the Unicode website, by references to specific sets of contributory
data files.

ISO/IEC 10646 expansion is by the publication of amendments, once they
have gone through international ballotting.

The two encoding committees, UTC and SC2/WG2, work hard to have significant
synchronization points, where the two standards can be matched up with
identical repertoires, but in between major versions it can get a little
tricky to match "versions" of the two standards.

The recent major synchronization point was Unicode 3.0. Here is the
information about that and subsequent synchronizations:

Unicode Version                   10646 Publication

Unicode 3.0                       ISO/IEC 10646-1:2000 (Second Edition)

=======================================================================

Unicode 3.1                       ISO/IEC 10646-1:2000 (Second Edition)
                                  + 2 characters from Amendment 1 to Part 1
                                  ISO/IEC 10646-2:2001

=======================================================================

Unicode 3.2                       ISO/IEC 10646-1:2000 (Second Edition)
                                  Amendment 1 to 10646-1:2000 (published 2002)
                                  ISO/IEC 10646-2:2001

=======================================================================

And prospectively, we anticipate the following synchronization in
repertoires sometime late next year:

Unicode 4.0                       ISO/IEC 10646-1:2000 (Second Edition)
                                  Amendment 1 to 10646-1:2000 (published 2002)
                                  Amendment 2 to 10646-1:2000 (to appear 2003?)
                                  ISO/IEC 10646-2:2001
                                  Amendment 1 to 10646-1:2001 (to appear 2003?)

But this situation is somewhat muddied by the fact that SC2/WG2 is also
starting on a project to merge Part 1 and Part 2 of 10646 into a single
standard, in which case it is conceivable that in 2003 or 2004 the coalesced
10646 will be reissued, presumably as ISO/IEC 10646:2003 (or 2004) (Third Edition),
rolling in the amendments. If so, we would return to a relatively simpler
publication situation, as for Unicode 3.0.

--Ken

> Then, Why can't we see "utf8-3.1" or "utf8-3.2" for Unicode 3.1 and 3.2 recpectively?
>