I was just browsing through the ActiveResource code and I wonder why
the use of .xml throughout on the URLs? Wouldn't it make more sense to
just have ActiveResource set the content type to application/xml and
then use paths without the extensions?
In general, I think we're moving away from the idea of using the Accept
header as the primary way of determining the content type. The
extension makes a lot of things simpler. Including page caching,
scanning logs (you can easily tell HTML requests apart from XML ones),
and off-hand exploration (curl http://example.com/people.xml).
So given that direction, I want to make sure that Active Resource is
encouraging that behavior.