Re: Collection paging section usefulness

>
>writing general purpose search endpoints is really complex. There is  
>one being developed by the W3C called SPARQL. Combine Atom, AtomOwl  
>and SPARQL and you get
>
>http://intertwingly.net/wiki/pie/PaceSparqlLink
>

Why does it have to be complex?  Let's not call it a "search endpoint", come 
to think of it, let's call it a "query endpoint".  Whatever language you wish 
to define as supported, becomes supported at whatever locations you choose to 
define (somehow).  Like so:

http://example.com/query?q={whatever-floats-yer-boat}&query-language='sparql'
http://example.com/2006/jul/query?q={...}&query-language='xquery'

And so on and so forth.  Because the one thing REST interfaces to query 
languages all seem to have in common, is the query goes between the curly 
braces.  All a client needs is a service document to tell them what the query 
URI is for a collection (somehow).

-EJB

>
>Henry
>
>Foaf name: http://bblfish.net/people/henry/card#me
>
>
>> -EJB
>>
>>>
>>> - Sylvain
>
>