W3C home > Mailing lists > Public > w3c-sgml-wg@w3.org > April 1997

Re: draft proposal for catalog resolution

From: Martin Bryan <mtbryan@sgml.u-net.com>
Date: Tue, 01 Apr 1997 18:41:11 +0100
Message-Id: <1.5.4.32.19970401174111.006908c8@mail.u-net.com>
To: W3C SGML Working Group <w3c-sgml-wg@w3.org>
Is there any real need for XML public ids to be SGML public ids?

I know I'm risking being heretical here, but it occurs to me that there can
be no strictly logical reason why XML public IDs should not conform to the
character set restrictions of URNs (as extended in the Mainz Unicode/HTML
Internationalization proposals) rather than those of SGML. Any SGML public
identifier would form a valid entry according to the URN rules. It is likely
that SGML will have to be revised to allow most 10646 characters at some
stage in the future (though when is anyone's guess). People using systems
which have an SGML parser in would be restricted to using the existing SGML
restrictions on character sets until such time as SGML changes, but those
simply using XML could take advantage of the advances in URN code set being
proposed if they so wish. Don't we get the best of both worlds
(Internationalized HTTP and old-fashioned file identification techniques) is
we adopt both approaches rather than going for the lowest common
denominator, the DOS 8/3 file naming conventions, that constrained the
original pi character set?
----
Martin Bryan, The SGML Centre, Churchdown, Glos. GL3 2PU, UK 
Phone/Fax: +44 1452 714029   WWW home page: http://www.sgml.u-net.com/
Received on Tuesday, 1 April 1997 12:43:39 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 10:04:23 EDT