Re: yet another revision of the charter

On Jul 12, 11:27pm, Roy Fielding wrote:

> Although it makes perfect sense to have a generic URI (UR(L|N))
> syntax, it does not follow that a generic URC syntax is useful,
> and what exactly would a generic URN resolution procedure be?

A generic resolution procedure might talk about what to try first,
what to use as a fallback, how errors should be reported, how
access controls might work, etc.

> A generic mechanism for identifying URCs as URCs would be good,
> but should be done immediately.
> Identifying suitable sets of characteristics for any URC, such that
> they support URN resolution, would also be good.  Is that what you
> mean by a generic syntax?

No, what I meant by a generic URC syntax was how we should set up
the general scheme for people to create custom URCs. There are lots
of different sorts of info that people will want to jam into URCs
to build their special applications. How do we let them do so
while still providing a reasonable measure of interoperability?

This was the big impetus behind the SGML URC draft put out by
Terry Allen and myself. According to that spec, URCs do not have
to be sent in text/sgml (although they can). Instead, novel URCs
have a SGML DTD to define their elements.

> I don't think that the WG should develop any of these.  I think
> it should continue to act as a forum for critiquing the system
> builders' specifications, perhaps analogous to a media type
> advisory group [Larry, did the IETF ever do anything about that?].

If these syntaxes are to be standardized, who else should do it?

Ron Daniel Jr.                email:
Advanced Computing Lab        voice: (505) 665-0597
MS B-287  TA-3  Bldg. 2011      fax: (505) 665-4939
Los Alamos National Lab
Los Alamos, NM,  87545    tautology: "Conformity is very popular"

Received on Thursday, 13 July 1995 11:01:12 UTC