http://user at example.com -- It's already there

Boris Erdmann boris.erdmann at googlemail.com
Sat Oct 21 09:00:16 UTC 2006


I don't get the point of the thread, because

handling of  "user(at)example.com" which ist to be
interpreted as "http://user(at)example.com" already works
perfectly as of today e.g. with the current JanRain
implementation (at least the consumer side).

What does this have to do with the spec anyway(?):

a) http://user(at)example.com  is a perfectly valid URL
b) URL delegation providers can choose to support discovery
    for such an URL
c) An IdP can choose to support discovery for such URLs


So what is the discussion about?

a) confused users?

    If an IdP choses to support "user(at)example.com"
    they will try to avoid presenting the "http://user(at)example.com"
    string to their users and use "user(at)example.com" instead

    Its up to the IdP to clarify with the user, that at some
    RPs their ID might be presented as "http://user(at)example.com"


b) implementation?

    Well, if the spec doesn't exclude use of "http://user(at)example.com"
    then everything is fine, and if properly implemented already works as of
    today.

The only thing that has to be done, is:
The Provider hosting the regarding "http://user(at)example.com" URL
has to check for a "user@" prefix when being contacted by HTTP at
example.com.


Regards,

Boris Erdmann



More information about the specs mailing list