[Home] [By Thread] [By Date] [Recent Entries]

  • From: Tim Bray <tbray@t...>
  • To: xml-dev@l...
  • Date: Tue, 09 Oct 2001 13:24:47 -0700

At 09:48 AM 09/10/01 -0400, Champion, Mike wrote:
>I haven't seen much response to Adam Bosworth's commentary on the state of
>XML and the W3C:
>http://www.xmlmag.com/upload/free/features/xml/2001/11nov01/et0111/et0111.as
>p

For those who don't know, Adam's this real smart guy who led
QuattroPro dev for Borland, then went to Microsoft and led the
development of Access and IE4.  Now he's at BEA.

I thought the interesting thing that Adam suggested was that
maybe the big mistake with XMLSchema is that it's trying to
be all things to all factions in the XML community.  He 
claims that it's really lousy for the purposes of the 
XML-as-datagram-message-interchange people - which is his
main focus.

Maybe the notion of one central overarching schema language
just doesn't pay off at the end of the day.  We need a bunch
of smaller special-purpose content constraint facilities
tuned to particular application classes.  You could, with
only a moderate leap of the imagination, draw a parallel
to Unix - there's a common view of all data as a sequence
of byte-accessible file primitives, then a large family of
specialized tools that do useful things with the data and
a good clean mechanism for combining them.  [That's the
historical view - later-generation tools like perl, emacs, 
and web browsers conflated all sorts of different data 
processing functions in a useful way, but they had a lot 
of experience to build on].  

Hmm... what's the XML equivalent of "|"? -Tim


Site Map | Privacy Policy | Terms of Use | Trademarks
Free Stylus Studio XML Training:
W3C Member