Goals of XML-DEV (was Re: IDL?)

[Peter Murray-Rust]

At 19:11 28/12/97 +0100, Matthew Gertner wrote:
>Having silently followed the discussion on this list for many weeks, I am
>delighted to see the latest initiative to produce a parser backend
>specification which is both simple and elegant. I have also been following
>the DOM activity very closely, and I very much agree with David on this.
>There needs to be a clear distinction between what the goals of this list
>are (besides a lot of very stimulating discussion) and what the DOM is
>doing. 

Thanks for your support, Matthew. This initiative is very central to what I
see the aims of XML-DEV are.  After nearly a year it may be worth some
gentle discussion of those goals.

<PREAMBLE>
XML-SIG, XML-WG, XML-DOM, etc. are all *formal* parts of the W3C process.
The W3C is composed of a large number of member organisations who pay
non-negligible subscriptions to support its work. XML-DEV is *not* part of
the W3C process and I would see it as secondary - but hopefully
complementary - to any formal W3C process.  Thus if the W3C already had an
initiative to produce a simple parser backend, I would recommend that it
should not be pursued on XML-DEV. Similarly, if any activity on W3C appeals
to the W3C it would be entirely appropriate for them to subsume some or all
of it.
</PREAMBLE>

XML-DEV is "a list for XML developers".  Henry Rzepa maintains the list
membership and mechanics and I act as unofficial "moderator" - no messages
are formally moderated.  I would see the following as outside the remit of
XML-DEV:
	- development *of* XML(XLL/XSL) [the W3C groups are for this]
	- non-XML markup languages or information management systems
[comp.text.sgml or appropriate newsgroup]
	- FAQs for XML [http://www.ucc.ie/xml] and answering beginners' queries
[comp.text.sgml]
	- current awareness for XML [http://www.sil.org/sgml/xml.html]

I would very much like XML-DEV to be responsible for creating XML resources
of value to the developer community (and possibly beyond that).  A resource
is something tangible that can be used by more than one person. Examples
(all of which have occurred on XML-DEV) could be:
	- XML-compliant software
	- SGML DTDs converted to XML
	- protocols for dealing with whitespace, characters, etc.
	- interfaces for software
	- entity sets in XML
	- clarification of how to *implement* the specs
	- test documents (including "torture" tests)
	- collections of useful resources (software, documents, etc.)
	- know-how about management of distributed documents (XML with MIME, jars,
etc.)
	- performance considerations
	
I cannot speak for the WG or SIG, but it is clear that the formal processes
are very busy "just" constructing the language specs. They do not have
resources to create testbed systems and it has been my hope that resources
made available on XML-DEV have been of value in confirming (or questioning)
the *implementability* of the spec. For example, when the first generation
of parsers came to be written, it became clear that the initial spec for
parameter entities caused problems in interpretation and implementation;
the WG subsequently revised PEs to a simpler form. XML-DEV is a place where
early implementers can announce their resources, ask for volunteers to test
them, give feedback to the WG/SIG. 

One of my main motivations is to try to limit unnecessary "semantic drift"
in XML implementations. Although the syntax of XML is very carefully
defined it semantics are (?deliberately) not.  Different implementers will
use their experience (or lack of) with SGML to interpret the spec in
different ways. this flexibility can be valuable, but it should be clear
that it is happening. Here is an example:

<!DOCTYPE FOO SYSTEM "foo.dtd" [
]>
<FOO>bar</FOO>

What would you expect a parser to do with such input? What would happen if
the file/URL "foo.dtd" couldn't be found? Already the parser writers have a
variety of *implied* semantics for this document. What I would hope is that
we didn't evolve an early *inconsistent* set of behaviour for this sort of
thing so that we end up with software-dependent documents. This is a real
danger with XML if there is no public analysis of implementations.

I believe that XML will need an RFC-like area for semantic and related
problems - I suggested the use of a namespace "XDEV" for this. If this is
provided elsewhere by a formal body, fine. Until such time, XML-DEV is
available for communal systematisation if sufficient people feel it
worthwhile. In a virtual environment you can never *force* people to do
things they don't want to but I and others will hope to nurture any growing
points we see :-)

	P.
Peter Murray-Rust, Director Virtual School of Molecular Sciences, domestic
net connection
VSMS http://www.nottingham.ac.uk/vsms, Virtual Hyperglossary
http://www.venus.co.uk/vhg

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)