Re: Concern with URI normalization text; proposed change

Roy T. Fielding wrote:

> On Thursday, February 19, 2004, at 03:36  PM, Graham Klyne wrote:
> 
>> Further to my earlier message [1], I've discussed the issue of URI 
>> normalization with some colleagues and we'd like to propose the 
>> following small change of wording with respect to [2].
>>
>> ...
>>
>> Section 6.1, para 2, final sentence:
>>
>> The suggested change is to this sentence:
>> [[
>> Therefore, comparison methods are designed to minimize false negatives 
>> while strictly avoiding false positives.
>> ]]
>>
>> To be:
>> [[
>> Therefore, comparison methods are designed to minimize false negatives 
>> while strictly avoiding false positives when used for purposes of 
>> retrieval.
>> ]]
>>
>> Rationale:
>>
>> This reinforces the earlier comment that "URI comparison is performed 
>> in respect to some particular purpose" [section 6 intro], and I think 
>> provides the necessary get-out for those purposes other than retrieval 
>> for which the normalization processes described can result in false 
>> URI-equivalence (i.e. in circumstances where existing applications may 
>> legitimately deliver differing results).
> 
> 
> Umm, no.  Aside from being difficult to understand due to the trailing
> qualifier, it is factually incorrect.  URI comparison has nothing
> to do with retrieval.  False negatives are false regardless of purpose.
> The purpose being discussed before that is the goal for which the
> cost/benefit trade-off is balanced, which could be different for each
> of a hundred different types of retrieval.
> 
> ....Roy
> 

Notwithstanding worries about the phrase
"when used for purposes of retrieval."
without any such qualifier the sentence is false.

For example, if a URI http:/example.org/namespace is used as an XML 
Namespace the similar URIs HTTP:/example.org/namespace or
http:/example.org:80/namespace are not names for the same XML namespace.

Thus techniques on the comparison ladder other than byte-for-byte 
comparison do not work for comparing namespace URIs when used for the 
purpose of naming namespaces. This counterexample (and others, for example 
from RDF), indicates that the sentence is false.

I believe I discussed Graham's proposed modification with him, and I felt 
that it attempts to allude to these issues without making too big a deal of 
it. Some change is necessary.

Jeremy

Received on Tuesday, 6 April 2004 05:15:23 UTC