[OpenID] OpenId Association Timeout Recommendations

Recordon, David drecordon at verisign.com
Fri Feb 9 19:48:44 UTC 2007


Have wording you'd like to propose?

--David 

-----Original Message-----
From: Granqvist, Hans 
Sent: Friday, February 09, 2007 8:51 AM
To: Recordon, David; 'David Fuelling'
Cc: 'security at openid.net'; 'general at openid.net'
Subject: RE: [OpenID] OpenId Association Timeout Recommendations

Current draft only details the value of trust roots, not encryption
strength. 

What's 'correct' for one may not be 'correct' for another. Security
assumptions rarely make a good foundation for specs. 

If the protocol is subject to such a huge hole it should be spelled out.

Hans

 -----Original Message-----
From: 	Recordon, David
Sent:	Thursday, February 08, 2007 11:38 PM Pacific Standard Time
To:	Granqvist, Hans; David Fuelling
Cc:	security at openid.net; general at openid.net
Subject:	RE: [OpenID] OpenId Association Timeout Recommendations

I don't think it is a reasonable assumption to make that people are
going to be running SSL with a NULL cipher suite in these situations.  I
think the spec is quite clear in the fact that you need to do TLS/SSL
right in order for it to matter.

So yes, there are MITM attacks if you're on an untrusted network and not
correctly using TLS/SSL.

--David

-----Original Message-----
From: general-bounces at openid.net [mailto:general-bounces at openid.net] On
Behalf Of Granqvist, Hans
Sent: Thursday, February 08, 2007 10:29 AM
To: David Fuelling
Cc: security at openid.net; general at openid.net
Subject: Re: [OpenID] OpenId Association Timeout Recommendations

> However, the spec seems to indicate that if SSL/TLS is used, then 
> Direct Verification is ok (Section 15.1.2, first line of 2nd 
> paragraph).  Do you agree?

In principle, yes, I do.  But SSL is such an ephemeral notion.
For instance, you can run SSL with NULL cipher suites so that traffic
goes in the clear.

To me, it seems that a RP that knows how to properly set up and use SSL
to verify the OP (with PKI trust processing) would probably want to
equally properly OpenID-associate.

The original intent of DV was for usage scenarios ("ajax") where proper
SSL is not normally nor easily available nor implementable. 

-Hans


_______________________________________________
general mailing list
general at openid.net
http://openid.net/mailman/listinfo/general



More information about the general mailing list