March 21, 2007

OpenID Negates the Need for P2P SIP

I just now read a post by Alex Mayrhofer where he suggests that OpenID can be used as the identifier in a P2P SIP network, which till now used a centralized id server. Apparently he is going to float this idea in the P2PSIP working group meeting. Since he is coming close to an idea that I had proposed almost 4 years back, I think I am going to make it public: one does not need OpenID to realize the ultimate p2psip (intentional change in the case).

The problem with the proposed P2PSIP is that it requires a critical mass of subscribers and that some of them have to become supernodes to route further traffic. If only two people want to communicate with each other and without requiring any body elseís assistance P2PSIp will not help you: in this respect it is no different than Skype. But P2PSIP and now Alex are missing a simple observation. I can store the IP address and the port number on which my SIP client will listen for incoming calls in a web page and my partner can visit that page before initiating a session request. This simple directory service almost eliminates the need for SIP proxies. Of course the problem is not fully solved because of the pesky NAT/FW traversal. If I am being served by a UPnP enabled router then the SIP client can establish the mapping rule; otherwise the web server can be used as the relaying point without requiring any SIP capabilities at all.

But with OpenID one can achieve something more. The web site can now provide the mapped IP address and port number by taking into account the OpenID of the initiator, just like Relevance Engine from iotum will do. What I mean is that the resultant mapping will be my SIP client for my family members and that of a voice mail server for business contacts, if the request is made on a weekend. I hope that after reading this Dean changes his mind regarding his remark about SIP not requiring the services of OpenID.

I am posting this in a hurry so as to lay a claim to this proposal that I had developed during the summer of 2003. Later on I presented this to my current employer during the interview (to show off my prowess) and have subsequently approached a few angels. Since I was hoping to realize in an implementation I did not make it public. Today my hand is forced to make it public and coincidentally a realized implementation will be available soon enough.

Posted by aswath at March 21, 2007 03:52 PM
Related Posts Widget for Blogs by LinkWithin
If you do not have an OpenID, then please use



I can vouch for the fact that Aswath has been floating this idea for quite a while. He's a voice in the wilderness which is usually right.

Posted by: Tom Evslin at March 21, 2007 08:55 PM

It's slowly dawning on me that OpenID is extremely disruptive, as it lets the user point to a directory service of arbitrary complexity, which can aggregate multiple relationships, and only the user is in control. "Circles of trust" is all about bizdev. The funny thing is, telcos could be OpenID providers, and it could end up proving extremely popular and profitable - if then can figure it out. They could insinuate themselves into all kinds of 3rd party applications -- but because the users want them involved, not because of their market power in access.

Posted by: Martin Geddes at March 22, 2007 11:51 AM

Martin, I am glad you have changed your mind about the nature of OpenID and that it is the user who has to decide to give business and presence in the access business does not assure success. We sure could use your articulation on this topic.

I still need to work on convincing you that there is no money in it. :-) Only then OpenID is truly disruptive.

Posted by: Aswath at March 22, 2007 04:00 PM

Copyright © 2003-2014 Moca Educational Products.