RE: [search-ws-comment] searchRetrieve Accept Parameters

[Matthew Dovey <m.dovey@jisc.ac.uk>]

> Don't get me wrong, I'm supportive of urls that give some indication of the
> format - and most implementors will likely expose different formats via
> different URI patterns (its a little easier to dispatch on anyway), the question
> is only whether that should be a part of the spec or not - and I suggest not.

Some of the client applications built using SRU have been very thin clients consisting of little more than client side XSLT generating a HTML interface. It is because of these use cases, that there was pressure to ensure that there was a mechanism within the spec to achieve all the functionality via URLs

Matthew