Researched this a bit more. Restful Web Services book suggests when
access foo.html, specify a CONTENT-LOCATION header that points to the
canonical representation of the resource:
i.e.
GET /foo.html
response:
content-location: /foo
<body>
Stephan Koops wrote:
> But I think it's the very very wide used kind to add the file type as
> extension. This is also an easy way to allow it to be used by the
> browser. The other accept properties (charset and encoding) are
> technical things, what the user does not interest. But the file format
> is interesting for the user, and also the language. Sometimes a user
> want to override, what the default is in the browser.
>
> Also the resource is identified by "/foo", independent of the media type
> (or extension) and the langage. So I think it is also useful to put the
> resource identifier in the front and the details of retrieving to the end.
>
> BTW, I think, HTTP servers, e.g. apache, supports this feature vice
> versa: request a file, e.g. "index.html", in english, than the apache
> look for file index.html.en (or index.en.html?)
>
> Stephan
>
> Bill Burke schrieb:
>> What I mean is isn't the convention:
>>
>> /html/en/us/foo
>>
>> rather than:
>>
>> /foo.en.html???
>>
>> In the spirit of REST, (reusing existing formats/styles), I vote the
>> spec follows whatever the convention is rather than making up its own.
>>
>> Bill
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe_at_jsr311.dev.java.net
> For additional commands, e-mail: dev-help_at_jsr311.dev.java.net
>
--
Bill Burke
JBoss, a division of Red Hat
http://bill.burkecentral.com