Private Use proposals

From: Doug Ewell (
Date: Tue May 21 2002 - 11:51:59 EDT

William, thank you for responding positively to my pre-proposal for
"lock" and "unlock" symbols, whatever they may be called.

I do feel I need to comment in regard to your two messages, totaling 19
KB, which are largely focused on the Private Use Area and quasi-official
codifications of its usage.

First, almost any idea, good or bad, can be summarized in much less than
9,000 bytes. It would really help not only the list but also your own
cause if you could post a brief summary to the list with a link to your
Web site, where the full proposal could be posted in all its glory. I'm
sure I'm not the only one who is tuning out after the first 2 KB or so,
possibly missing some good ideas buried in the rubble.

Many of the issues relating to the PUA have been discussed numerous
times on this list. We all know, as you state in one of your posts,
that Unicode is committed to leaving the PUA free and available to all
users, to the point that they will not sanction any "semi-official"
mappings of characters to the PUA nor any indexing mechanism to
reference such mappings. I used to wonder why there was no link on the
Unicode Web site to ConScript, since it seemed to me like a creative use
of Unicode. Now I understand that such a link might be misinterpreted
as an official endorsement of ConScript.

I am still getting the sense that certain other Unicode concepts are not
being understood:

1. There is *strong* opposition to encoding additional presentation
forms for alphabetic characters. Ligatures are presentation forms.
Beginning with version 3.1, Unicode has stated that alphabetic ligatures
may be formed with the help of U+200D ZERO WIDTH JOINER, or
automatically by the font without explicit encoding. (As Michael
Everson pointed out in his "zero-width ligator" papers, the
automatic-formation approach requires hairy contextual analysis in cases
such as Fraktur.) But the whole reason for inventing these solutions is
that additional Latin ligatures are EXTREMELY unlikely to be encoded.

2. There is no evidence that encoding a character in the PUA increases
its likelihood of being accepted into Unicode. For most of the
characters and scripts that have been added in the last 5 years or so,
it would be difficult or impossible to find a publicly accessible PUA
encoding which helped get the character or script encoded. (Deseret and
Shavian were encoded in ConScript; whether that helped get them into
Unicode or not, I don't know.) It is certainly not the case, as I have
said before, that PUA encoding is any sort of guarantee or fast-track to
UTC and WG2 acceptance.

2a. In particular, the specific PUA code point given to a character has
NO bearing whatsoever on the likelihood that the character will be
"promoted" to Unicode, or on which code point it will be assigned if
that does happen. It is complete fallacy to state that "the idea behind
suggesting that U+E707 to U+E70F be used is that promotion to U+FB07
through to U+FB0F would be as straightforward as possible." There is no
relationship. Deseret, for example, was laid out two different ways in
its ConScript PUA encoding and its final Unicode encoding.

3. Finally, the bottom line about the PUA is this: Free love. Do your
own thing, baby. Right on. If you want others to share your encoding,
whether it involves Latin ligatures or colors or italics or the Theban
or Utopian alphabet or whatever, then publish the encoding on your Web
site. Groovy. But don't expect that action to have any bearing on what
UTC or WG2 does. They want formal proposals, and they have an official
form. If I decide there is enough support for "lock" and "unlock" to
warrant a proposal (the grass-roots vote so far is 3 for, 1 against), I
will fill out the form. Doing a PUA implementation is fine, but has
nothing to do with formal proposals.

Sorry for chewing up so much additional bandwidth.

-Doug Ewell
 Fullerton, California

This archive was generated by hypermail 2.1.2 : Tue May 21 2002 - 12:34:24 EDT