W3C home > Mailing lists > Public > www-tag@w3.org > October 2003

RE: Action item on syntax-based interoperability

From: Elliotte Rusty Harold <elharo@metalab.unc.edu>
Date: Thu, 23 Oct 2003 22:26:09 -0400
Message-Id: <p06002001bbbe3b230de3@[192.168.254.4]>
To: "Dare Obasanjo" <dareo@microsoft.com>, "Tim Bray" <tbray@textuality.com>, <www-tag@w3.org>

At 12:32 PM -0700 10/23/03, Dare Obasanjo wrote:
>I see, instead of countering or agreeing with my point you throw an 
>anti-Microsoft slur instead. Either way, the point still stands that 
>web technologies like CSS are based on shared data models not just 
>shared syntax. Syntax based interop is a great start but it isn't 
>the end all and be all of interoperability on the Web.
>

Mostly I just want to make really sure no one confuses XOM with DOM. 
But when Microsoft accuses me of embracing and extending a standard, 
all I can think is, "Hello, pot? This is the kettle calling. You're 
black."

But on the point, I don't think DOM and CSS are interoperable; and 
for a long time I did blame Microsoft's atrocious implementations of 
these specs for sorry state. But maybe it's really not Microsoft's 
fault. Maybe no one could correctly implement data model 
interchangeability. Maybe the problem is the fundamental flaw of 
trying to exchange a data model rather than syntax. Why have things 
like XML-RPC and SOAP succeeded where more data-model driven 
technologies like CORBA, RMI, and DCOM failed? Why have even more 
syntax-oriented systems such as REST and RSS done even better than 
half-breeds like SOAP and XML-RPC? Why are the syntax based systems 
so much easier to implement? Why does it seem the further a system 
gets from syntax, the worse it works?

I'm not a theoretical proponent of syntax like Walter Perry. I'm an 
experimentalist and an observer. I look out in the world, and I see 
syntax-based systems working far better and coming into production so 
much more easily than the data model systems. I think at this point 
in time the burden is on the data model proponents to demonstrate 
that their systems are interoperable and implementable. They 
certainly haven't proven that with DOM, CSS, and XSLT, If these are 
the best examples you've got of interoperable data models, then there 
are real problems.

-- 

   Elliotte Rusty Harold
   elharo@metalab.unc.edu
   Processing XML with Java (Addison-Wesley, 2002)
   http://www.cafeconleche.org/books/xmljava
   http://www.amazon.com/exec/obidos/ISBN%3D0201771861/cafeaulaitA
Received on Thursday, 23 October 2003 22:43:16 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:22 GMT