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

Re: How did "public identifier" get its name...

• To: xml-dev@l...
• Subject: Re: How did "public identifier" get its name...
• From: ari@c... (K. Ari Krupnikov)
• Date: 31 Jul 2003 13:25:46 -0700
• In-reply-to: <20030731153122.GU9926@s...>
• References: <15725CF6AFE2F34DB8A5B4770B7334EE022DC706@h...><20030731153122.GU9926@s...>
• Sender: ari@l...
• User-agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.3

John Cowan <jcowan@r...> writes:

> Bullard, Claude L (Len) scripsit:
> 
> > The problem of the Universal Identifier concept is that 
> > it assumes the web is the universe and vice versa.  It 
> > builds unreliability into the system.  Definitions that 
> > include the term 'universal information space' are silly.
> 
> IMHO it would have been better to decree that in XML the public id,
> rather than the system id, must be a URI.

FPIs have much richer structure than typical URIs, certainly URLs:
language, encoding and file type (DTD, ENT, DOC, etc.). All these are
available in MIME headers, XML declarations, xml:lang headers... When
you GET the entity over HTTP and the author used xml:lang.

Ari.

• References:
  • RE: How did "public identifier" get its name...
    • From: "Bullard, Claude L (Len)" <clbullar@i...>
  • Re: How did "public identifier" get its name...
    • From: John Cowan <jcowan@r...>

• Prev by Date: RE: more politics
• Next by Date: Re: more politics
• Previous by thread: Re: How did "public identifier" get its name...
• Next by thread: RE: How did "public identifier" get its name...
• Index(es):
  • Date
  • Thread