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

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

It just isn't possible to be even a single order of magnitude faster in
the general case.  An alternate encoding for XML can provide gains in 2
ways: file size, or parser CPU usage.  I've never seen a format which
provides significant file size gains without high CPU cost (although you
can sometimes play games and push all the CPU cost to the writer, so
that the reader is cheap, for example).  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.

Note the key constraint in my opening sentence: "in the general case".
Some scenarios clearly benefit from an alternate encoding, but different
scenarios demand different trade-offs.  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. 

-derek

> -----Original Message-----
> From: DuCharme, Bob (LNG-CHO) [mailto:bob.ducharme@lexisnexis.com]
> Sent: Monday, November 22, 2004 12:49 PM
> To: 'Robin Berjon'; Michael Champion
> Cc: xml-dev@lists.xml.org
> Subject: RE: [xml-dev] Hostility to "binary XML" (was Re: [xml-dev]
XML
> 2004 weblog items?)
> 
> Robin Berjon wrote:
> 
> >Michael Champion wrote:
> >> The 'binary XML' stuff got a lot less hostile reception than I
> >> expected.  Is the world ready to hear that XML 1.x text
serialization
> >> is not suitable for wireless applications, is this old news, or
what?
> 
> >during a recent presentation I tried to outrage the audience with
> >some over-the-top pro-binary XML positions (expecting to use the
> >push-back to moderate them)
> 
> And in Mike's talk, he did not do this. He put everything I had heard
in
> context, including counter-arguments to the various arguments, so
there
> was
> little reason to get hostile. For example, there was his point
(whether
> his
> own or quoting of other people) that just because XML doesn't suit a
> particular class of applications well doesn't necessarily mean that
XML
> should be twisted inside out to accommodate that class of apps;
perhaps it
> means that that class of apps should look elsewhere for its needs.
> 
> His answer to my question at the end convinced me to maintain my
current
> level of skepticism: apparently all of the talk of alternative XML
> encodings
> being much more efficient than text XML are based on consistent use of
> document classes for each test, so this consistency means the kind of
> redundancy that makes compression much easier.
> 
> When someone prototypes an encoding that is orders of magnitude more
> efficient for arbitrary XML, which Mike said that no one had done yet,
> I'll
> more seriously consider the possibility that a binary XML standard
might
> be
> worth the trouble.
> 
> (Mike, please correct me if I misremember. And it was nice to finally
meet
> you!)
> 
> Bob DuCharme   www.snee.com/bob       <bob@
> snee.com> weblog on linking-related topics:
> http://www.oreillynet.com/pub/au/1191
> 
> -----------------------------------------------------------------
> The xml-dev list is sponsored by XML.org <http://www.xml.org>, an
> initiative of OASIS <http://www.oasis-open.org>
> 
> The list archives are at http://lists.xml.org/archives/xml-dev/
> 
> To subscribe or unsubscribe from this list use the subscription
> manager: <http://www.oasis-open.org/mlmanage/index.php>