- From: Larry Masinter <masinter@attlabs.att.com>
- Date: Thu, 15 Jun 2000 09:26:34 -0700
- To: "James Clark" <jjc@JCLARK.COM>, "Henrik Frystyk Nielsen" <frystyk@microsoft.com>
- Cc: "David Turner" <dturner@microsoft.com>, <XML-uri@w3.org>, "Andrew Layman" <andrewl@microsoft.com>
I think the group should reconsider compatibility with RFC 2557's algorithm
for comparison.
1) URIs are absolutized, but unless there is an obvious, embedded
base, the base 'thismessage:/' is used.
2) After this, the results are compared octet-by-octet, without any
translation, decoding, equivalence.
# This comparison will only succeed if the two URIs
# are identical. This means that if one of the two URIs to be
# compared was a fictitious absolute URI with the base
# "thismessage:/", the other must also be such a fictitious
# absolute URI, and not resolvable to a real absolute URI.
In order to get the effect you want as far as namespace name comparison,
all that's necessary is to be careful about the definition of the 'base'
for namespace names that are in relative form. I would posit that the
URI of the containing document is irrelevant, and that the namespace name's
base should either be the namespace name of the container (i.e., relative
URIs as namespace names are relative to the containing namespace name),
or else a fictitious base.
Using 'thismessage:/' solves the problem of having to deal with silly
relative references that start with '..': they're invalid.
Larry
Received on Thursday, 15 June 2000 12:28:03 UTC