Re: Comments on URIs, URLs, and URNs

"Ian B. Jacobs" wrote:
> 
> Hello,
> 
> I printed and read the URI clarification document [1] that
> was published today. I have some questions about section 1 "URI
> partitioning". I'm not a URI expert, so please forgive
> what may be naive questions. However, as a naive reader, I
> was expecting this type of information from this document.
> 
> If I have understood correctly, we are asked (in adopting the
> contemporary view) to dispense with the necessary classification
> of a URI scheme as either a URL scheme or a URN scheme.
> Instead, we should think of URI schemes on their own, and we
> MAY classify certain of them (e.g., based on properties they share).
> URI and URN are two examples of URI schemes that have been
> classified.

Er... no, that doesn't look right. I don't think I understand
what you wrote, in fact.


> What is not clear to me from the document is why I should
> be interested in this new perspective. What can I do with my
> newfound independence from the one-or-the-other view?

You can stop worring about whether, say,

	http://www.w3.org/1999/xhtml
or
	urn:pin:1
or
	glorp:lsjfliw3jli

are a names or an addresses. You're free to use them as either.
You can observe, syntactically, that they're URIs (cf RFC2396).
You can also see, from inspection, that the second one
is a URN, since it starts with "urn:..." and that the other
two or not.

And you can go to the URI scheme registry
to find that the community has not yet agreed to start
using the glorp: URI scheme.

In the URI scheme registry, you can look
under "http:" to see that www.w3.org should
be looked up in DNS, and under
"urn:" to see that pin: should be looked
up in the URN namespace registry.



> Is it
> simply that I should stop thinking about URLs and URNs
> all the time in association with URIs, and this will allow me
> to come up with other interesting properties of some schemes)?

If you're writing specs, yes, you should stop thinking
about URLs.

You can keep thinking of URNs, but realize URN is
not some fuzzy concept; it means exactly: a
URI that starts with "urn:...", whose semantics is governed
by the specs you'll find registered in the URI scheme
registry under "urn:".


> What other properties have people found useful in practice?
> Sections 1.1 and 1.2 only talk about URNs and URLs, so
> I didn't get a sense of other interesting directions.
> 
> Why did people move from the classical view to the contemporary
> view?

Umm... how about this:

because it makes all the distinctions that are necessary
to get the various jobs done. There are more distinctions
folks used to make (in the classical view) but these
distinctions are (a) controversial, and (b) not necessary
to Getting Stuff Done.

> What caused the change in viewpoint?
> 
> When I read the title "1.3 Confusion", I was expecting a
> brief discussion of technical confusion - sources of
> confusion that might make people have one view or another
> or mix the two (or others). Instead, the section focuses
> on inconsistencies among (or within) documents. Perhaps
> the title should be adjusted so that readers don't
> expect to find "Common sources of confusion around URIs."

Hmm... I'm not sure what to make of this... I'll have
to study it some more.

> I suspect that the latter topic will be explored in more
> depth in a W3C Activity on URIs.
> 
> Thanks,
> 
>  - Ian
> 
> [1] http://www.w3.org/TR/2001/NOTE-uri-clarification-20010921/
> 
> --
> Ian Jacobs (ij@w3.org)   http://www.w3.org/People/Jacobs
> Tel:                     +1 718 260-9447

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

Received on Tuesday, 2 October 2001 18:22:46 UTC