Skype's Outage - Raising Questions About P2P

I was under the weather when all the Skype problems came up last week, and it's been done to death by now. One thing, though, I'd like to draw attention to is the topic of supernodes, which has been picked up by a number of other bloggers. This is a pretty technical aspect of the peer-to-peer architecture, and I really only understand it at a basic level.

However, that's enough to know that it's an important - and proprietary - part of what makes Skype work so efficiently. On the other hand, many believe it has a lot do with why Skype went down last week, and of course, raises all sorts of questions about the P2P model as a serious alternative to conventional telephony.

SightSpeed is a company I've followed for a while, and while their core offerings are similar to Skype - P2P text chat and video - they do not use supernodes in their architecture. The point here is that Skype is not the only - or definitive - model for P2P, and we all know that its reliance on proprietary technology for its secret sauce does not make everyone happy in the Voice 2.0 world.

I'm mentioning them here because their CEO, Peter Csathy, has a good blog, and this morning, he published a guest post from his CTO, Aron Rosenberg. I highly recommend it if you want a better understanding of the pros/cons of supernodes, and in contrast, how SightSpeed relies instead on standards and SIP-based elements for their architecture. Aron also explains the advantages of their model, especially for business/enterprise applications, with the implication being that what happened to Skype would not have happened - and did not happen - to SightSpeed. Read it and tell me what you think. I just want to make sure their voices get heard - it's an important message.


Technorati tags: , , ,