Re: A Potentially Useful Property - Last Informative Proposal

You may start by the "Unicode roadmap" page which summarizes the
current proposals.

I think it is enough, and it does not have to become part of the UCD,
not even in an informative part. Exatly because neither the name, nor
the codepoint is fixed (and there are even proposals for which not
even the base block has been determined, the proposed code points are
listed like as in "U+xxx01", and will change several times due to
reorderings, merges, or splits of the proposed characters, because of
the unification process, or removed/delayed due to lack of information
and distinct priorities).

What you'll want is a specialized search engine to look into current
proposals. However not all proposal documents are published (there are
lots of beta documents, or documents with contradictory opinions,
between which nothing has been decided). So the search would given
links to documents with very variable status, including those that
have been rejected or abandonned by their authors.

2012/7/28 Richard Wordingham <richard.wordingham_at_ntlworld.com>:
> As writing systems can be quite complex, there is often useful
> information about a character that cannot be found in the Unicode
> Character Database or the Unicode Standard, but is present in one of
> the proposals leading up to its inclusion.
>
> To this end I have half-jokingly suggested that it would be useful to
> have an informative numeric property 'last informative proposal'. A
> non-missing value xxxx would indicate that the last proposal to contain
> useful information about it would be held as
> http://std.dkuug.dk/jtc1/sc2/wg2/docs/nxxxx.pdf .
>
> Now, there can be problems with establishing the value - it may be
> require a judgement call as to when revised proposals started to
> provide less useful information. There are also subtle variations in
> the URI. Now, I don't seriously expect the Unicode Consortium to take
> on the task of setting up such a database, but are there any publicly
> available databases with such information or a reasonable
> approximation? A poor approximation - but a lot better than nothing -
> is the derived age property which records when a character entered
> Unicode. One problem with using a search engine is that the name and
> code point proposed for a character can change during the encoding
> process.
Received on Fri Jul 27 2012 - 22:06:31 CDT