Hi Sandford,

Sorry, when I said the server returns a different response I wanted to
say that the server acts different, thus, doing different (and faster)
processing than with a GET request. As an example, a GET request would
return an object from the server while a HEAD request would only check
if the object exists. The data transferred to the client on a HEAD
request will be just the headers, as it should.

When you say XHR objects do not support HEAD, is this true for all
cases or only for IE6? We no longer support IE6, so that is not a
concern for us.

Thank you,

Eneko

On Sun, May 16, 2010 at 12:18 PM, Sanford Whiteman
<[email protected]> wrote:
>> We are currently developing a new REST API for our proprietary CMS and
>> it will have support for HEAD requests, this is, as you mentioned, the
>> server wont process the full request as if it was a GET but return a
>> different response instead.
>
> Hmm,  but  it  shouldn't  return a different response. The "spirit" of
> HEAD  (it  is  an RFC SHOULD) is that it is identical to a GET for the
> same entity, except that only the headers are returned.
>
> That's  why  it's more of a convenience method that's most useful when
> you don't control the server and you want to make sure that it doesn't
> put  extra  bytes  on  the wire. If you do control the server (as with
> your  proprietary platform) there's no advantage over sending a custom
> request for /EN/TI/TY/HEAD /EN/TI/TY/FULL etc.
>
> I  agree that it would be nice if Moo supported it (although every XHR
> object  doesn't -- I think this would break IE 6 BC), but it shouldn't
> be  a  stopping  point  for  your project, since you can wrap the same
> functionality in the request parameters.
>
> -- S.
>
>



-- 
Eneko Alonso
Software Engineer
http://enekoalonso.com
San Luis Obispo, California

Reply via email to