to Home posted 10/05/2011 in Short and sweet about 3 minutes read

Skype the hype

When the acquisition of Skype by Microsoft was announced, I checked some of the early posts from various sources that tried to research why did Skype manage to grow so much and became simple too big to ignore or avoid. Apart from the various “free, easy, cute” descriptions, the one that remained in my mind was “p2p”. At first I thought that was a typo. But then it made more sense than anything else.

When Shawn Fanning introduced Napster to the masses, the p2p or peer-to-peer buzzword immediately unleashed the mindset where “servers” lost their importance, and “clients” become important and integral parts of the network. Skype built its entire network on that simple guideline, and precisely “p2p” made the network virtually limitless when it comes to scalability. Many services failed and were buried when the number of users grew more than the servers could handle, and the scalability issue becomes a dominant force when startups start attracting masses.

One of the negative sides of Skype’s infrastructure was that it could turn regular clients into “supernodes” that could route traffic to unreachable clients. While some went as far as calling it “stealing bandwidth”, that concept made Skype popular in business environments where proxies and firewalls prevent stuff from communicating with the web. It is all in the details. The fantastic work of Skype engineers that managed to keep Skype extremely non-centralized is to be hailed and presented over and over again as a fantastic recipe for dealing with exponential growth platforms. P2P while being the demon ISP’s around the world fear the most (torrents eat bandwidth), is actually the savior of modern networks. In a p2p network, you would stream a YouTube video from someone inside your ISP network (if available) instead from YouTube servers, making the entire process a) faster, b) cheaper. And precisely YouTube is the #1 “legal” bandwidth consumer in the world.

Peer-to-peer is very underestimated. Mainly because of the negative usage it has flourished under. One of the biggest values of P2P networks is that the more peers it has, the stronger it is. Few know that the root DNS system the World Wide Web relies on is a variant of P2P, that’s why it is so hard to bring it down as no single source has everything under control.

Having spent so much words on P2P, I can’t avoid referencing projects that smell like Skype, but miss the key ingredient that made Skype an multibillion $8.5B success., a fantastic idea planted in Slovenia targeted all the problems Skype has. No software installs, no proxy headaches, clean and simple web interface. Everything very nice on the outside, but the core is not P2P (update below), and in an industry (telecommunications) where bandwidth = quality = cost, there is a virtual limit such a platform can grow to. After that, there is no sky – just a roof.

— update —

And I love being wrong and proven wrong. Since Flash 10.1, P2P VoIP is integrated into Flash. So what I wrote above about is partially wrong (there is still one brain that connects the two dots). The roof I mentioned is much much higher, with the only problem remaining that Adobe holds the core of the service.

Skype on the other end will now become a business division of Microsoft, and there are already brainstorming sessions on the power a Microsoft + Nokia + Skype bridge has. While I’m convinced Microsoft splashed the cash mainly to stay in the web game (a game they just can’t get things right), the idea of the triad working together and using all available synergy is simply mind blowing. Mobile operators, you are warned.

~ the end ~

to Home posted 10/05/2011 in Short and sweet


blog comments powered by Disqus

More reading