[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

RE: the infoset is two infosets (or even three?) [was: Re: [xm

• To: 'Tim Bray' <tbray@t...>, xml-dev@l...
• Subject: RE: the infoset is two infosets (or even three?) [was: Re: linking, 80/20]
• From: "Bullard, Claude L (Len)" <clbullar@i...>
• Date: Thu, 29 Aug 2002 13:06:42 -0500

They are confusing.  There is probably a healthy consensus on that. 
Leaving aside some technical changes, why are they confusing?

Part of that originates in the use of URIs with HTTP.  The problem 
for the w3c is that that is architectural, so, comes with the turf. 
If a URL is URI and identity is primary not emergent, then the 
definitions as given will confuse anyone not hardcore mathematician.

Part of that is an insistence on the part of some vocal advocates 
that URN resolution software not being widely deployed should not 
be used in namespace values.  A tautology, and again, confusing.

The insistence by some that they are MerelyDisambiguators when 
empirical testing shows they are semantic loading devices is 
confusing.  Worse, it leads to expectations not proven correct 
in production.  Change the namespace and the code fails that 
worked before the change.  Somewhere a bit changed state. That 
is not syntax.

It is confusing.  

Ummm... so XML-Dev has to figure this out before the W3C listens?  Huh?
If that is all we have to do, then what do we need the W3C for?

Great.  Aaron, care to propose the elimination of the default? 
Any others?  We can vote and get right to changing the code. 
No waiting.

len


From: Tim Bray [mailto:tbray@t...]

John Cowan wrote:
> Aaron Skonnard scripsit:
> 
>>So is there any chance that someone from the W3C is listening so that
>>these endless debates can be transformed into something constructive? 
> 
> 
> Well, yes, I'm a member of the Core WG

And I'm on the TAG.  There are occasions when the message coming out of 
xml-dev is fairly coherent, and when that occurs it tends to get noticed 
(examples would be SAX, RDDL, RESTifying SOAP) and not just at the W3C. 
  An example of something that has emerged and is starting to get a bit 
of traction is the fairly widespread distaste for XML Schema.

As John notes, there is nothing remotely approaching xml-dev consensus, 
however rough, on issues such as namespaces and the infoset.  I think 
that this is a fairly accurate reflection of the state of affairs out 
there in our profession. -Tim

• Follow-Ups:
  • Re: the infoset is two infosets (or even three?) [was: Re: linking, 80/20]
    • From: Paul Prescod <paul@p...>

• Prev by Date: Re: the infoset is two infosets (or even three?) [was:Re: linking, 80/20]
• Next by Date: RE: Linking global context
• Previous by thread: Re: the infoset is two infosets (or even three?) [was:
• Next by thread: Re: the infoset is two infosets (or even three?) [was: Re: linking, 80/20]
• Index(es):
  • Date
  • Thread