[Ohta]
: ISO 10646 is done. OK. It is a complete standard of "characters" with
: its own definition of a "character". But it is incomplete as an text
: encoding standard.
[Adams]
: Since neither 10646 nor Unicode purport to be a "complete text encoding
: standard", you are amiss in attributing this as a deficiency.
[Ohta]
: The problem is that ISO 10646/Unicode does not address the issue for
: displaying or comparing strings which contains more than a single
: "character".
Tell me, since when did any (international or national) character set
define either display or sorting behaviour? Does JIS X 0208 define
sorting order? Does ASCII? Do ASMO 449 (Arabic) or ISCII (Indian)
standards define display behaviour for what are clearly complex scripts?
The answer is no. None of them do. The same answer applies to 10646:
it should not define either. They are in the application domain. But,
and this is an important but, should there be additional standards that
assist in defining in this behaviour? Probably. That's precisely why
both SC2 & SC18 have undertaken a joint project to develop a Character/Glyph
model that assists in defining display behaviour. Other SCs are also
developing standards for defining sorting behaviour. Although interrelated,
such ancilliary work needs to be separated from the character set standard
itself, whose task is *only* to enumerate a set of characters assigned to
code points according to some encoding scheme, *and no more*.
[Adams]
: 10646 and Unicode are the foundation upon which one can indeed build a
: complete text encoding standard (where text is understood as content,
: structure, and appearance).
[Ohta]
: I designed ICODE exactly for such a purpose. How do you think about it?
I don't know anything about it. Perhaps you could send me a paper describing
it. [Metis Technology, Inc., 522 Atlantic Ave., Boston, MA 02210]. I would
be interested in looking at it from an academic perspective, but not from
any practical perspective since it won't go anywhere as a real standard.
[Adams]
: However, good engineering practice dictates modularity, not
: monolithicness. 10646 and Unicode achieve the first important step
: in the process towards this goal. To accuse them of failing to
: solve all the problems of text encoding in general is a little bit
: naive don't you think?
[Ohta]
: But that was the job we expected to SC2.
Who are "we"? I believe it was Mark Twain who said "never say 'we' unless
you are royalty or have tapeworm." Perhaps *you* expected it, but you seemed
to neglect to inform SC2 that that is what their charter should have been.
If you would bother to research the history of the charter of SC2/WG2 or of
SC2 in general you would find that your expectations wildly contradict
reality.
Regards,
Glenn Adams