Subject: Re: [xsl] Bug in Schema for XSLT 2.0?|
From: Dimitre Novatchev <dnovatchev@xxxxxxxxx>
Date: Wed, 16 Feb 2005 21:55:33 +1100
On Wed, 16 Feb 2005 10:34:43 +0100, Peter Gerstbach <peter@xxxxxxxxxxxx> wrote: > Quoting Dimitre Novatchev <dnovatchev@xxxxxxxxx>: > > > So why not have two schemas from the XSLT 2.0 WG: the current one with > > all bells and whistles and a much simpler one that will be easily and > > immediately dropable in most of today's not so clever schema-oriented > > text editors? > > What is the intended purpose of the schema in current XSLT 2.0 Draft? Is it > validation of stylesheets? Or is it providing a simple way to implement entry > helpers for XML editors? Or feeding as an input for source generators > (XML Data > Binding)? > > There are numerous products which are great although the Schema support is not > complete. Complete support is good of course, but mostly not the main focus of > development. > > In my opinion a much simpler schema would be great. But it would be > difficult to > decide what to leave out. But the thing with complexContent inside a > SimpleType > might be a good candidat. ;) The fact that there are different schemas for the same language -- one more complex and the other simpler -- does not at all mean that the simpler schema has "left out" some parts of the language. To be more specific, Norm Walsh has produced an RNG schema for XSLT 2.0 (and it can be trang-ed to XSD) without cutting any XSLT features from it. http://norman.walsh.name/2004/07/25/xslt20 Here's a small quote from this same blog entry: "Having a grammar for 2.0 is nice (a usable grammar, I mean, the W3C XML Schema version in the spec doesn't do me any good)" I agree that we are speaking about *usable* grammars. As another note, if an xml schema is based on the meaning of the super-confusing sentence quoted by Michael Kay, it is only natural to observe that the majority (if not all) schema-based xml editors will fall in the category of "buggy products" with "undocumented limitations". Now, I know that Norm Walsh is one of the WG members and I greatly admire his work. However, why shouldn't it be possible to have a great and useful schema as his is, but coming *officially* from the XSLT 2.0 WG? Cheers, Dimitre Novatchev.