From: Kenneth Whistler (kenw@sybase.com)
Date: Thu Jul 17 2003 - 14:59:37 EDT
William spilled another ocean of digital ink. Found bobbing
in that ocean was the comment:
> >Roozbeh and I assigned two unencoded characters for Afghanistan to
> >the PUA, and we encourage implementors to use them until such time as
> >the characters are encoded.
>
> Yes. ... Now that at least one of them has been approved for
> encoding by the Unicode Technical Committee there is now a long period of
> waiting during which Private Use Area encoded data can be produced. This
> does seem unfortunate and for individual symbols such as these I would hope
> that the people who are in charge of Standards might like to consider asking
> if the United Nations and the World Trade Organization could perhaps arrange
> for some faster way of achieving agreement.
*rolls eyes*
It seems rather unlikely that getting the United Nations and the
World Trade Organization involved in trying to amend JTC1 standards
directives would be a recipe for speeding anything up. :-)
> It does seem so very slow for
> the twenty-first century with so many electronics communications facilities!
> Why does legacy data have to build up and resolving the problem take so long
> for just a few symbols?
Because amending and updating a standard is effectively the same
task whether it involves 1 additional character or 181 additional
characters. There are a large number of stages, approvals, reviews,
and other tasks involved -- which are there for a reason, to ensure
the stability and orderly maintenance of the standard.
> I would have thought that with a reasonable
> infrastructure that those two code points could have been formally added
> into regular Unicode and ISO within a couple of weeks.
The whole idea of adding a couple code points this week and then a
couple more next week, and then another next month, and so on is,
well, just nuts. It would destroy effective version control and
would create a situation where implementers were unsure just what
was in the standard and when it would change further. It would
*damage* the standard rather than improve anything.
A character encoding standard is not just a laundry-list registration
of characters that people happen to notice this week. As such, it
is not advisable to create a mechanism whereby new characters
are noticed, approved, and "registered" on a weekly basis.
> An ocean of digital ink! I like that phrase.
As well as producing the oceans, clearly.
> That person added that people
> have been telling me for a long time that PUA codes are not suitable for
> interchange.
Not suitable for *public* interchange, because, by definition, in
public interchange the receiver will not be a party to whatever
*private* agreement defines their usage, and so will not be able
to interpret them.
>
> That puzzles me, because I thought that it was alright to interchange
> Private Use Area codes if there is an agreement as to their meaning in a
> particular situation.
Yes, a *private* agreement for *private* interchange. That, as Michael
tried to tell you, is why we call them *private* use characters.
> Also, Unicode 3.0 mentions the possibility of
> publication of Private Use Area assignments
Anyone is free to publish anything they wish, including lists of PUA
assignments.
> in the section on the Private
> Use Area.
But the Unicode Consortium will not publish such lists in the
Unicode Standard or on its website in any official way.
>
> So what is the official position please?
I just stated it. If you want chapter and verse:
"All code points in the blocks of private-use characters in the
Unicode Standard are permanently designated for private use--no
assignment to a particular, standard set of characters will ever
be endorsed or documented by the Unicode Consortium for any of
these code points."
-- The Unicode Standard, Version 4.0,
Section 15.7, "Private-Use Characters",
p. 398, 2003 [forthcoming]
> This is important to me because I
> have been proceeding in the belief that suggesting three Private Use Area
> code points for use in interactive television systems is entirely proper and
> compliant with Unicode and the ISO standard.
It is. But other participants on this email list have been telling
you that they are not interested in your *particular* use of
private use characters.
--Ken
This archive was generated by hypermail 2.1.5 : Thu Jul 17 2003 - 15:55:03 EDT