RE: using namespaces to version

["Bullard, Claude L (Len)" <clbullar@ingr.com>]

No, but consider that the "systems" work 
well enough.  The definitions don't.  We 
wave our hands over them and then wink 
while a resolver is written or someone 
sharp implements a catalog.  What is 
RDDL really?  It's a catalog XML-Dev 
built so a namespace reference could 
be resolved a year after "reasonable 
minds" blessed non-resolution while 
"experienced minds" sighed and said, 
"that won't hold".  They declare a 
minimal victory, confuse the hell out 
of the world, then come back a year 
later, wave their hands over it and 
say, "RDDL me this."  

It's voodoo engineering.

Version information can be collected 
outside the document or inside.  If 
it is collected inside, it should use a 
Public reference to name the type and 
as system reference to get an instance 
of the type.   What we have now is a 
religion of string monism in the 
definitions but MIME outside.  Practical 
but every three months, it has to 
be explained all over again.

Len 
http://www.mp3.com/LenBullard

Ekam sat.h, Vipraah bahudhaa vadanti.
Daamyata. Datta. Dayadhvam.h


-----Original Message-----
From: Warren Hedley [mailto:w.hedley@auckland.ac.nz]
Sent: Thursday, May 03, 2001 6:40 PM
To: xml-dev@lists.xml.org
Cc: Norman Walsh; Bullard, Claude L (Len)
Subject: Re: using namespaces to version


"Bullard, Claude L (Len)" wrote:
> 
> It is probably a mistake to name the process of
> change inside any component participating in that
> process.

Fair enough, but hard to implement at the document level
given current filesystems and MIME types for HTTP.

Or have I missed your point completely?

To be more specific, the topmost level that I can see me
being able to specify the language and version for an
XML document is the DOCTYPE declaration. Is this a good
solution?

-- 
Warren Hedley
The Bioengineering Research Group
The University of Auckland
New Zealand