Reflections on Spring VON 2006 – Peer-to-peer SIP

Henning Schulzrinne and Henry Sinnreich were championing P2P SIP at VON. These men are legendary and I respect tham a great deal. That said, the P2P SIP efforts leave me a bit unfulfilled. Chasing the likes of Skype is beneath such men, and yet that’s exactly what P2P SIP feels like.

The Internet pioneers didn’t copy some existing commercial implementation of on-line systems, like Compuserve or Prodigy. They leap-frogged such systems, thinking “outside the box” as the popular saying goes.

Before Vonage, nobody was talking about ‘broandband phone service’ in that form. Afterwards, Vonage-style VoIP was all anyone could see and nobody was talking about something like Skype. Now we have Skype, and again, we’re acting like it’s the end, rather than what it really is, an evolutionary, closed, proprietary, step along the way — a step that will pass, along with all the others before it.

So instead of copying Skype, we have to be thinking well beyond it. Some colleagues and I did SIPshare some 18 months ago, to demonstrate the application of SIP in pure peer-to-peer (serverless) environments — P2P SIP is old news.

Ebay’s multi-billion dollar Skype acquisition seems to have created an industry wide temporary blindness. Trust me, the wealth that will be created by the things that will follow Skype will pale that number. That’s where we need to be putting our enegry and that’s where we need great minds like Henning and Henry.

3 comments for “Reflections on Spring VON 2006 – Peer-to-peer SIP

  1. P2P SIP is not about "Skype envy", though I must confess I love Skype for showing how well P2P works. P2P SIP is aimed at using the Internet for how it was initially designed: End-to-end use of the "dumb network" by clever applications.
    The present attempts by the telephone company style voice industry to load the net with an ever heavier VoIP infrastructure (see for example IMS and TISPAN) gives more urgency to the need to have P2P as an alternative to the rebirth of the failed IN, AIN, ISDN, BISDN, etc.
    It would be a pity to let users pay for all this bloated, legacy minded VoIP infrastrucure that serves only the cash flow of VoIP vendors and possibly also to keep competing traffic away from bundled Internet+VoIP networks. The bloated IMS-TISPAN will not do any good to service providers either, since P2P SIP will save them from the present VoIP infrastrucure complexity and cost.
    Last but not least, putting multimedia communications back at the edge of the net will support innovation, without the need to rebuild the VoIP network in the middle for new services and applications.

  2. As you know Henry, I’m all about the edge :-). We can say P2P is not "Skype envy" all we want, but when it walks like a duck…

    What I mean is, whenever many (most?) people talk about P2P SIP they seem to pretty much focus on Skype parity.

    And I also agree that nobody wants the likes of IMS, et al. The only thing they’re good for is sucking up lot’s of VC money.

    Do you know of the JXTA effort? That has been out for a long time and is no longer just about Java: http://www.jxta.org/ It seems like the P2P SIP efforts are duplicating a lot of work already done within JXTA and perhaps there is a way to bring the two together in some way.

  3. I keep on reading about how bad IMS is… and I cannot stop thinking about how short of "good" my experience using Skype is.

    When I use Skype, I need my computer, I need some QoS settings in my network (that I usually dont have) and it is difficult and challenging if I need mix and match other non-Skype applications on the go.

    Skype is great because it is free. But because it is free (or almost free to call PSTN), I hardly can use it during the rush hour.

    Perhaps it is my problem or my ISP’s problem… But for the time being, when we talk about opening a system, about setting proper Call Admission Control and QoS in the network, about billing and about opening/closing/mixing, different applications… There I think that the complexity of IMS/TISPAN adds value.

Comments are closed.