W3C home > Mailing lists > Public > w3c-dist-auth@w3.org > October to December 2003

RE: How to use DTDs, or not to (was: RE: ACL and lockdiscovery)

From: Lisa Dusseault <lisa@xythos.com>
Date: Sat, 18 Oct 2003 00:06:27 -0700
To: "'Julian Reschke'" <julian.reschke@gmx.de>, <dennis.hamilton@acm.org>, "'Geoffrey M Clemm'" <geoffrey.clemm@us.ibm.com>, <w3c-dist-auth@w3.org>
Message-ID: <001f01c39546$596b27f0$f8cb90c6@lisalap>


> > So does this mean we should simply remove the DTD-style 
> definition of 
> > the 'prop' element (as well as 'owner' and 'resourcetype') from the 
> > spec?  I believe the natural-language specification should be 
> > sufficient to define these anyway.
> 
> The drawback would be that this would break the DTD's syntax. 
> An alternative is to keep ANY, and make sure that the 
> description says what this actually means.
> 
It wouldn't break the DTD syntax if we only use DTD fragments to 
formally define those elements that can be formally defined.  I thought
the idea of omitting the full combined DTD appendix was a generally
acceptable idea, as long as the DTD fragments were still there
for most of the elements.

Lisa
Received on Friday, 17 October 2003 15:07:24 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:44:05 GMT