<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Eric Norman wrote:
<blockquote cite="mid05057082887becff0e681edc0e9bb533@doit.wisc.edu"
type="cite">
<pre wrap="">On Dec 6, 2007, at 11:47 AM, Dominick Accattato wrote:
</pre>
<blockquote type="cite">
<pre wrap="">What happens when an OpenId provider is down:
<a class="moz-txt-link-freetext" href="http://www.alexanderinteractive.com/blog/2007/09/disadvantage-of">http://www.alexanderinteractive.com/blog/2007/09/disadvantage-of</a>-
openid-and-web-services.html
</pre>
</blockquote>
<pre wrap=""><!---->
I'll just mention that this problem isn't peculiar to OpenID.
Informantion cards have the same problem. With the current
design, every use of a managed card requires contact with the
IdP. Similarly for Shibboleth.
The current Shibboleth answer is to build in enough redundancy
such that it never happens (barring natural catastrophes).
Regardless, it's is a question that needs an answer.
Eric Norman
</pre>
</blockquote>
At IIW, there were several discussions around 'synonyms' in various
contexts. The idea being that RPs can record multiple OpenIDs per
account, perhaps in some cases discovered semi-automatically through
AX, and thus have the ability for users to ride out N-1 simultaneous OP
failures or account closures. <br>
</body>
</html>