[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: XML Public Indentifier
- From: mrossi@csc.com
- To: xml-dev@lists.xml.org
- Date: Wed, 05 Sep 2001 12:58:23 -0400
Why would you even try to dereference a name when it doesn't reference
anything in the first place? Everyone seems to quick to forget that N is
for Name, L is for Locator. If you're need is to identify something a name
(URN) usually does the trick, if your need is to locate/dereference
something a locator (URL) seems awefully well-suited. The real question is
whether you wish to identify or locate to begin with.
My $0.02, I like the theoretical purism of URNs for namespaces becasue
they're just supposed to be names. But that doesn't seem to get you all the
way there in practice. The convenience of being able to retrieve a schema
of some sort (or other supporting information) for a namespace simply by
using the name assigned to it (if it's a URL) is undeniable, as
demonstrated by RDDL.
As for Public Identifiers, a URN there would seem quite nice as an
alternative to FPIs. Maybe it would even be mapped to a URL in a catalog
file. :-)
Michael A. Rossi
Computer Sciences Corporation
mailto:mrossi@csc.com
856-983-4400 x4911
Tim Bray
<tbray@textua To: xml-dev@lists.xml.org
lity.com> cc:
Subject: Re: XML Public Indentifier
09/05/01
12:27 PM
At 08:15 AM 05/09/01 -0700, David Brownell wrote:
>A better solution is to use only URNs as system IDs in XML, using
>some URN resolution mechanism.
There's a chicken-and-egg problem here. I am reluctant to start
stuffing URNs in XML until there's a good chance that they can
be dereferenced and used to retrieve things. At the moment, my
own computer contains no URN dereferencing software that I'm aware
of. On the other hand, I don't see a big chance that URN
resolution software is going to become ubiquitous until there's
a killer app for URNs; I suspect XML entity retrieval isn't it.
So I'd say we're stuck with URLs for the moment. Using URNs
in the PUBLIC ID slot seems plausible though. -Tim