- From: James Clark <jjc@jclark.com>
- Date: Sun, 07 Nov 1999 11:36:55 +0700
- To: Larry Masinter <masinter@parc.xerox.com>
- CC: xsl-editors@w3.org, Makoto MURATA <murata.makoto@fujixerox.co.jp>, Ted Smith <Tsmith@parc.xerox.com>, Tim Berners-Lee <timbl@w3.org>
It's not clear to me that the "Associating Stylesheets" Recommendation is in fact in error. Although I agree that fragments don't have MIME types, I don't think the use of a type pseudo-attribute on the processing instruction when the href pseudo attribute includes a fragment identifier is necessarily inconsistent with that. I see two alternative interpretations which could be consisent: - the type pseudo-attribute is giving the MIME type of the complete resource not the fragment - the type pseudo-attribute is using a MIME type as a convenient identifier of the stylesheet language, just as the STYLE element in HTML uses a type attribute to specify the stylesheet language of the STYLE element (even though elements don't have MIME types) I therefore think the appropriate course of action is to leave the example as it is and in due course take up in the XML Core WG the issue of whether there is an error in the "Associating Stylesheets" Recommendation and, if so, how to fix it. Larry Masinter wrote: > > OK, so, you can't change the example without fixing the > "Associating Stylesheets" recommendation. > > I think there are three possible courses, then: > > a) fix the "Associated Stylesheets" recommendation first, > before issuing the XSLT recommendation. > > b) change the XSLT recommendation to add a note that the > example is compliant with "Associating Stylesheets" > but likely to be wrong and revised. > > c) issue this recommendation, and then (immediately?) > issue errata revising both of them > > I prefer (a). > > Larry > -- > http://www.parc.xerox.com/masinter
Received on Sunday, 7 November 1999 02:10:59 UTC