Re: Resolving IRIs for resource documents in SVG Tiny 1.2

> This seems to imply that if I have a document that contains many resource
> references with the same relative URI, I have to perform an HTTP transaction
> for each one to see if any of them result in a redirect. Wouldn't it make a
> lot more sense to use the pre-redirect URI as the dictionary key, so I can
> avoid a lot of network transactions?

Although at a first glance I was tempted to agree with you, after a
more through analysis I believe the behavior proposed by the
specification is probably the more appropriate. In a stateless,
non-persistent process, your approach seems to make more sense but, if
one uses an gateway-like URI which is redirected to a final location
depending on state, than this can become problematic. Note that this
is labeled "counterexample", not to be taken as an good/advised URI
usage pattern (at all)! :-)

Maybe the specification could advert that, depending on URI
organization and redirect usage, the proposed behavior could lead to
(potentially) useless network transactions. Nevertheless it seems to
be a somehow general URI organization matter - (un)RESTfull [1] and
related concepts. Linking to an external document on the subject could
help bringing focus to the matter without inserting much noise on the
specification.

Thoughts?

 Helder Magalhães

[1] http://en.wikipedia.org/wiki/Representational_State_Transfer