Nowhere does the spec define what constitutes equivalent
URI templates. Does
http://example.org/{a} == http://example.org/{b}
for example? Barring any further type annotations
to {a} and {b}, one would have to assume that is does.
An explicit rule would allow for validation where
URI templates are used, like WADL. This would
seem to be a pathological condition:
<application
xmlns="">
<resources base="http://example.org/">
<resource path="{a}">
<method href="#one"/>
</resource>
<resource path="{b}">
<method href="#two"/>
</resource>
<resources>
</application>
Given the URI templates implied, there's no
way to determine which method to dispatch to. With
a defined notion of URI template equivalency, a
validation rule could be defined for WADL processing.
- Jeff

IMH, URI Templates comparison should be case-sensitive and
character-by-character meaning that your two examples are not
equivalent. The URIs resulting from each MAY be equivalent, but that's
a separate question.
- James

Jeffrey Winter wrote:

Nowhere does the spec define what constitutes equivalent
URI templates. Does

http://example.org/{a} == http://example.org/{b}

for example? Barring any further type annotations
to {a} and {b}, one would have to assume that is does.

An explicit rule would allow for validation where
URI templates are used, like WADL. This would
seem to be a pathological condition:

<application
xmlns="">
<resources base="http://example.org/">
<resource path="{a}">
<method href="#one"/>
</resource>
<resource path="{b}">
<method href="#two"/>
</resource>
<resources>

</application>

Given the URI templates implied, there's no
way to determine which method to dispatch to. With
a defined notion of URI template equivalency, a
validation rule could be defined for WADL processing.
- Jeff