proposal for capabilities lookup

[Dan Libby]

On Friday 18 November 2005 20:24, Martin Atkins wrote:
> Ernst Johannes wrote:

> By providing a hint in that initial request, that list can be reduced by
> one item, which is valuable to server implementations built in to things
> such as weblog apps which would otherwise have to spend a lot of time
> generating an HTML document which is going to be immediately discarded.

Agreed. Many (most?) Identity pages will be dynamically generated. A GET 
request seems to me like mostly a waste of bandwidth and server CPU.  I would 
actually suggest that we make the following YADIS design goals:

1) Capabilities document is always requested/retrieved in a single request.

2) a GET of the Identity URL is never required.


Note: alternatives to GET of Identity URL that I'm aware of include:
 a) a HEAD request, with capabilities somehow encoded in response header(s)
      - not sure I've heard any real arguments against this yet.
 b) a GET request of another [magic] URL
      - not a popular idea, magic is bad.
 c) an OPTIONS request
      - Seems designed for this purpose. Works for webDAV, maybe for us too?

-- 
Dan Libby

http://videntity.org/
 - One identity to login with them all
 - Social networking between websites and blogs