RE: [xml-dev] Hostility to "binary XML" (was Re: [xml-dev] XML 2004 weblog items?)

["Derek Denny-Brown" <derekdb@microsoft.com>]

> From: Liam Quin [mailto:liam@w3.org]
> Sent: Monday, November 22, 2004 1:34 PM
> 
> On Mon, Nov 22, 2004 at 01:09:06PM -0800, Derek Denny-Brown wrote:
> > Most of the CPU cost of parsing is related to the abstract model
> > of XML, not the text parsing: Duplicate attribute detection,
> > character checking, namespace resolution/checking. Every binary-xml
> > implementation I have researched which improves CPU utilization does
> > so by skipping checks such as these. At that point you are no longer
> > talking about XML.
> 
> One can do validation in the writer and then plausibly skip the sort
of
> checks you mention in a reader, and still be talking about XML, even
> with today's textual interchange formats.

The problem with that is that it now becomes possible to load invalid
XML.  Yes, that means you are using a non-conformant writer, but when
dealing with data formats designed for interop, the receiver should be
conservative, which means checks.  If it is an error for there to be two
attributes with the same name, then it should either be impossible to
represent, or checked in the reader.  Anything else is asking for
trouble.  That doesn't mean that parsers can't provide options to turn
off expensive checks, just that they should be enabled by default for a
generic implementation.

> > I have yet to hear of any proposed solution which successfully
> > balances the different demands. I'm not sure it is possible, without
> > creating a homunculus.
> 
> Neither am I, which is why W3C has a Working Group to investiate
whether
> it might be possible, rather than a WG to implement a homunculus :-)

And I totally concur with others, that spending the time to really do
some serious requirements gathering is key, and am glad to see it
happening.

-derek