- From: Daniel Veillard <daniel@veillard.com>
- Date: Sat, 11 Sep 2004 14:09:00 +0200
- To: Dan Connolly <connolly@w3.org>
- Cc: daniel@veillard.com, Jonathan Marsh <jmarsh@microsoft.com>, public-xml-core-wg@w3.org, www-xml-xinclude-comments@w3.org
[ fixing the WG email in the Cc: ] On Fri, Sep 10, 2004 at 06:06:59PM -0500, Dan Connolly wrote: > On Fri, 2004-09-10 at 17:48, Daniel Veillard wrote: > > [...] A bad test suite is more dangerous than no test suite, and at > > this point dilution of already scarce efforts doesn't sound wise to me. > > I see your point... on the other hand, a Recommendation that isn't > thoroughly tested can do more harm than good too. I would just like to point out that thorough testing of the spec in isolation is probably sufficient to not do harm if said spec defines its scope clearly. To go back to XInclude I do think we have a good coverage of the spec in isolation now, and that XInclude being defined as an Infoset transformation, i.e. clearly scoped as a post processor parsing allows to use it in a variety of contexts without doubts about its own behaviour. Where doubts may arise is when you are pairing XInclude with other processing which have their own specific requirements. To me this raises back the question of a canonical XML processing order which is something which we debated but never came to any official answer (and this may as well be kept unanswered). Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel@veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ |
Received on Saturday, 11 September 2004 12:09:06 UTC