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

Re: defaults and equivalence, same-document references

From: Roy T. Fielding <fielding@gbiv.com>
Date: Fri, 13 Feb 2004 00:42:43 -0800
Cc: uri@w3.org
To: Mike Brown <mike@skew.org>
Message-Id: <974CC5A2-5E00-11D8-8468-000393753936@gbiv.com>

On Tuesday, February 10, 2004, at 09:20  PM, Mike Brown wrote:
> Section 4.4 says that same-document references are determined only by 
> URI
> identity (not equivalence).
>
> Therefore, I would assume that none of these examples are same-document
> references, even though I, personally, think that retrieval of a new
> representation in order to satisfy the reference would be undesirable 
> in every
> case:
>
> Base URI                       URI reference
> ===========================    ========================
> file:///x                      file://localhost/x#y
> http:///x                      http://localhost/x#y
> http://www.example.com:80/x    http://www.example.com/x#y
> http://www.example.com/x       http://www.EXAMPLE.com/x#y

It will depend on the implementation.  I have clarified that.

> I would like to see defaults, whether they be defined by schemes (e.g. 
> http's
> "80" for port and file's "localhost" for host) or by the URI spec 
> (e.g. the
> recent addition of "localhost" for host for all hierarchical-path 
> schemes), be
> mentioned in section 6, because they're almost certainly going to 
> affect
> comparisons for equivalence.

Done.

> I would also like confirmation as to whether the intent of section 4.4 
> is to
> define identical string comparison as the only way to determine 
> whether a
> reference is same-document, ruling out the possibility of equivalence 
> being a
> factor.

The intent is to match what applications do in practice.  I have made
it clear that normalization is an optional step that can't be assumed.

....Roy
Received on Friday, 13 February 2004 03:43:06 UTC

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