(no subject)

MR-Received: by mta REDMS1.MUAS; Relayed; Tue, 15 Apr 1997 16:14:33 +0000
MR-Received: by mta RE5; Relayed; Tue, 15 Apr 1997 16:13:40 +0000
MR-Received: by mta RITIG4; Relayed; Tue, 15 Apr 1997 16:13:25 +0000
Disclose-recipients: prohibited
Date: Tue, 15 Apr 1997 16:14:33 +0000 (GMT)
From: Misha Wolf <misha.wolf@reuters.com>
Subject: Re: Cougar (the next version of HTML) and i18n
In-reply-to:
 <9912121615041997/A84582/REDMS2/11B47C0C0C00@RITIG4.RIT.REUTERS.COM>
To: Dave Raggett <dsr@w3.org>, Dan Connolly <connolly@w3.org>,
 HTML Editor <www-html-editor@w3.org>
Message-id: <7333141615041997/A84622/REDMS2/11B47C0E2100*@MHS>
Autoforwarded: false
MIME-version: 1.0
Content-type: TEXT/PLAIN; CHARSET=ISO-8859-1
Content-transfer-encoding: QUOTED-PRINTABLE
Importance: normal
Priority: normal
UA-content-id: 11B47C0E2100
X400-MTS-identifier: [;7333141615041997/A84622/REDMS2]
Hop-count: 2

Cougar, the next version of HTML, has a number of problems in the i18=
n area.
The problems I have found include many areas of incompatibility with =
RFC 2070,
such as:

1.  The document character set is limited to plane 0 of Unicode (aka =
the=20
    BMP).

2.  The <HTML> tag does not permit the LANG attribute.

3.  Language tags are defined as those of RFC 1766 plus those of the=
=20
    Ethnologue.

All of the above highlight a process problem with Cougar.  The author=
s are
clearly both competent and well-intentioned, but are doing a very lar=
ge and
complex job and haven't got the time to be experts in everything.  Wh=
at is
needed is the involvement of experts in specific fields, in this case=
, i18n.=20
At least one of the authors of RFC 2070 (maybe Martin D=FCrst) should=
 be invited
to help incorporate RFC 2070 in Cougar.

Point 3. above highlights another problem.  Because of the complexity=
 and
interdependence of the various standards and protocols in use on the =
Internet,
it is essential that the development of these standards is carried ou=
t in a
modular fashion, allowing those who have an expertise in a particular=
 area to
have visibility of, and comment on, developments in that area.  If RF=
C 1766 is
to be replaced by a new standard, an Internet Draft should be circula=
ted to,
and discussed on, the relevant mailing lists.  An HTML spec is emphat=
ically
*not* the right place to modify the definition of language tags.  Tur=
ning to
the Ethologue proposal in particular, this must *not* be allowed to p=
roceed as
is, due to the widespread use of a different set of 3-letter language=
 codes in
the library community.  These codes are currently being voted on for =
inclusion
in ISO 639.  In principle, it would be possible to include the Ethnol=
ogue's
3-letter codes as well as these other 3-letter codes in a daughter-of=
-RFC 1766,
by splitting the name space (eg prefix Ethnologue codes with "e-", ju=
st as IANA
language codes are prefixed with "i-").  This is *not* to be taken as=
 an
endorsement, on my part, of one or other (or, for that mattter, eithe=
r) set of
3-letter codes.

We need to:

a) fix the process,

b) fix the individual problems.

Misha



===

Received on Tuesday, 15 April 1997 11:16:44 UTC