[OpenID] SREG 1.x attributes

[Martin Atkins]

Peter Williams wrote:
> Here is a compromise: make it clear that both solicited and unsolicited responses to an auth can bear a ax etension ... alongside the sreg).
> 
> Ax comes across as: ask for auth, get assertion (with sreg), now go back and ask for ax, get ax response (over association keys).
> 

Is that really what the AX spec describes? That was not my 
interpretation when I read it, but then I may have been reading it with 
SREG-tinted glasses.

If the AX spec really is suggesting that the attributes be exchanged in 
a separate transaction, then we should totally fix that. AX is supposed 
to be a superset of the functionality of SREG.