Re: Localizable sentences (from Re: Plain text (from Re: Avoidance variants))

From: Philippe Verdy <>
Date: Sun, 29 Mar 2015 00:46:46 +0100

2015-03-28 14:51 GMT+01:00 William_J_G Overington <

> Doug Ewell wrote as follows.
> > It's the same as one of the concerns I have with encoding localizable
> sentences as characters.
> There aren't 20 or 50 or 100 sentences that people might want localized,
> but crores of them.
> Well, that seems like an example of a status point moving along a path
> upon the surface of a cusp catastrophe manifold in a mathematical model of
> decision making produced as an application of catastrophe theory.
> One instant there is a desire by people for zero of them to be encoded,
> then the next instant there is the desire by people for crores of them to
> be encoded.

You don't need Unicode encoding to encode localizable sentences. There are
various ways to do it.

Notably using URNs (or URLs to URN resolvers) where you can create your
registry of entities to encode (use the encoding scheme you wish, you'll
probably use unique identifiers such as UUIDs or a digital hash such as
SHA1 of an initial sentence in any language) or XML paths (within some XML
DOM scheme). These URNs can then be resolved either as images/icons (in
various formats), or as plaintext (encoded in standard Unicode)., or rich
text (such as HTML), or as glyph identifiers in some custom symbolic fonts.

If you're not able to create this registry, don't even think that Unicode
will host it (unless part of it demonstrates its usefulness for inclusion
in CLDR by wishes of CLDR contributors and votes in the CLDR TC). Then only
later, may be, some parts of CLDR data (or of some external data in a
registry shared publicly and with a sizeable and open community of users
that wish to extend the open use of their conventional encoding, such as
stenographic notations, or chemistry symbols; or emoticons/emojis) may be
encoded into plain-text Unicode (such as mathematics symbols which carry
thir own stylistic requirements for their admissible glyphs, but more
importantly their identity as *maths* symbols distinct from standard
letters with too variable forms).

Unicode is definitely not the first place to do your homework : you
**first** need to create or find a community of interest and start
documenting your processes and conventions in your own collaboration
workspace and make sure it is open to collaboration and that this workspace
has a measurable usage (outside yourself only).

Absolutely nothing you constantly propose in this list has demonstrated
this minimum requirement. All what you propose recurrently here is then out
of topic.

Unicode mailing list
Received on Sat Mar 28 2015 - 18:48:11 CDT

This archive was generated by hypermail 2.2.0 : Sat Mar 28 2015 - 18:48:11 CDT