W3C home > Mailing lists > Public > uri@w3.org > February 2004

Re: Canonical Form of URIs "/" and IRIs

From: Roy T. Fielding <fielding@gbiv.com>
Date: Mon, 9 Feb 2004 18:08:14 -0800
Cc: <uri@w3.org>
To: "Jeremy Carroll" <jjc@hplb.hpl.hp.com>
Message-Id: <FC34C8E6-5B6D-11D8-A8FB-000393753936@gbiv.com>

On Thursday, December 4, 2003, at 09:40  AM, Jeremy Carroll wrote:
> I have two comments on section 6.3 of
>
> http://gbiv.com/protocols/uri/rev-2002/draft-fielding-uri-rfc2396bis- 
> 03.html
> #canonical-form
>
> 1) suggest add the following additional rule:
>
> + For URIs following the generic syntax produce an abs_path of "/" in
> preference to omitting the abs_path
>  (this might need an additional example earlier in the doc
> http://example.com
> vs http://example.com/ )

Done.

> 2) may it be helpful to explicitly extend the rule:
>
> "Only perform percent-escaping where it is essential. "
>
> to considerations where the URI is being transported in a way that
> anticipates IRIs e.g. XML system identifiers - and to discourage  
> %-escaping
> of non-ASCII chars in such circumstances.

No, because such situations do not actually involve URIs until after
the IRI normalization is completed.

> (I note that XML Namespaces 1.1 is explicit about this:
>
> "Because of the risk of confusion between IRIs that would be  
> equivalent if
> derefenced, the use of %-escaped characters in namespace names is  
> strongly
> discouraged."
>
> http://www.w3.org/TR/2003/PR-xml-names11-20031105/

I would (and do) strongly discourage the use of IRIs for namespaces.
Any identifier that is primarily intended for machine consumption should
be constructed for maximum interoperability.

....Roy
Received on Tuesday, 10 February 2004 18:59:15 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:25:07 UTC