A 12:16 98-09-17 -0700, email@example.com a écrit :
> Jony Rosenne writes:
>> I object to the strange pseudo names attached to the Hebrew
>> characters and some other characters, for example in item 115
>> "ISO-8859-8 ISO-IR-138 ISO_8859-8:1988 ISO_8859-8 HEBREW".
>> The ISO names, possibly in abbreviated form, should be used.
>Although I still would like the names to be human-readable,
>I'm losing hope that we can agree on what those names would be --
>although the Unicode/ISO 10646 descriptive names would be obvious
>candidates. That may be what you mean, Jony, when you suggest
>using "the ISO names."
Remember that there are *at least* two "ISO names" for any character for
which there is a French version of an ISO/IEC JTC1/SC2 standard, the
English and the French (and there are such names in other national
standards too in other languages, for example Swedish and even Gaelic, two
languages that I know of as real cases in point). Russian is still an
official language of ISO and there are probably some names in Russian too
which are official as they might have been used in some Russian versions of
International Standards although this has been more rare in the last
decade. There is no such thing as a unique name for a character in ISO
standards. And in natural language, except perhaps in artificial languages,
this is hard to achieve too, synonyms always exist (otherside we would have
some difficulty in making dictionaries, all refernces being inevitably
circular in human language, we must be quite intuitive aand brilliant in
our infancy, all of us who learned to speak, to even learn our mother
>Another candidate is to use the Unicode values (e.g., U005C,
>U005D, etc.). They're not human-readable, but they don't give
>an advantage to English speakers, and they only use the
>*invariant* characters from ISO 646. If an international
>registry must use unreadable names, I recommend using the
> -- Sandra
>Sandra Martin O'Donnell
Indeed. At least, since 1995 (first at the ISO/IEC JTC1/SC2 Plenary in
Helsinki in June, reconfirmed in facts more clearly at a meeting in Tokyo
the same year [easy for me to remember, it was held exactly one week after
the last Québec referendum on its future (; ]), SC2 recognized that there
is only one unique standardized *identifier* (current terminology to use)
for each character, its Uxxxx value in the UCS, independently of coding.
*Identifiers* are now -- clearly -- recognized as different from *names*.
*Identifiers* are primarily for machine usage and *names* for human usage.
Humans can use the names they want if they can anchor them with
standardized identifiers. This can also be done via intermediary,
non-standardized, identifiers, which could be mnemonic for a particular
usage community, also using the same standardized identifier anchor (see
also ISO/IEC 14755 -- Input methods to enter UCS characters -- which uses
standardized identifiers for input and which also gives guidelines on how
to deal with character names in the user's language (including mnemonic
So since 1995, even 8-bit ISO standards are being published with this
reference to the central catalog of characters, i.e. ISO/IEC 10646.
City of Québec, Québec (still Canada, for those who have a doubt [*])
* There is also a Quebec, Texas, a Quebec, North Carolina,
a Quebec, Virginia, a Quebec Junction, New Hampshire and...
a Quebec, Jamaica (this one near Montmorency [**], like in Québec,
Québec, the only one with a mandatory written accent!)
** The Duke of Kent, father of Queen Victoria, stayed for years both
in Québec, then capital of Canada, and in Montmorency, a few
kilometers away near the 90-meter famous Montmorency waterfall,
surrounded by women and eating good food and wines, before being
married and fathering perhaps the most puritan queen of all times...
... the normal conflict of generations, I guess... (;
I live at an equal distance of those two sites (: -- a paradise...
---- End of publicity on my beloved city of Québec ----
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:41 EDT