Anne van Kesteren wrote:
On Thu, 14 Feb 2008 00:36:05 +0100, Ian Hickson <[EMAIL PROTECTED]> wrote:
On Wed, 13 Feb 2008, John Panzer wrote:
Some of today's APIs like Flickr put authorization evidence into URL
query parameters for CSR. It's mildly bad to do this because such
things are more likely to get logged and sniffed than headers, and you
can't separate the resource URL from the authorization proof being
presented to use it, which would be useful in caching.
Also agreed. That's one of the reasons that XMLHttpRequest + Access
Control together let you set arbitrary extension headers.
This is currently not the case for XMLHttpRequest level 2. Based on
feedback from Mozilla only Accept and Accept-Language can be set for
cross-site requests.
(Aside: Surely Content-Type is allowed as well?)
This rules out the use of AtomPub's (RFC5023) If-Match: header on PUT
for optimistic concurrency control, and the Slug: header[1] on POSTs for
suggesting the URI to mint. The first is especially troublesome.
It also eliminates the ability to do cache control (except crudely by
salting the URL, which of course fills up caches with dead data). It
makes it impossible to use the common X-Method-Override work-around for
intermediaries which don't support things other than GET and POST. It
prevents the use of the Range: header to get a subset of a resource.
And of course it prevents the use of any custom X- header for any purpose.
-John
[1] http://tools.ietf.org/html/rfc5023#section-9.7
