Joost Abandons P2P – Or Not?

Joost has informed its users via email that it will discontinue the support of its desktop client today and instead completely concentrate on its new web site. This is a big step for a company that once aimed to revolutionize online video with P2P technology, and whose founders previously succeeded with P2P apps like Kazaa and Skype. But it’s way too early to declare the death of P2P video streaming, as some seem eager to do in light of Joost shifting course.

Not only are others far more successful with P2P video clients, but it looks like Joost may bring back some elements of its software sooner or later. This includes P2P distribution, but also other social and interactive features that made Joost’s software unique. Maybe we’ll have to hold off writing the obituaries for both Joost and P2P just a little longer.

Joost has obviously spent a lot of resources developing a working P2P infrastructure for its client. Initially, it had some scalability issues, but it worked fairly well later on. We asked Joost CEO Mike Volpi what’s going to happen to his company’s P2P technology now, and he was quick to point out that this effort is not going to be wasted. “As we find the right application for that technology, we will reintroduce it into a service,” he told us, without specifying further what application this might be.

One possible scenario would be live sports or music programming. P2P works best when many people access the same video simultaneously. In fact, Chinese P2P start-up PPLive was able to handle 1.6 million concurrent viewers for the opening ceremony of the Olympic games. Numbers like these tend to still cause hiccups for traditional CDNs, but a well-done P2P solution is able to scale up without any major issues.

That being said, Joost did run into some trouble when it tried to live-stream March Madness earlier this year. But hey, there’s always next year, right? Major sports events could also help Joost to overcome a big obstacle of its previous client-only model: Joost essentially forced users to install a dedicated video client — only to discover that the content wasn’t worth the effort. Video distribution companies like Move Networks have demonstrated that people don’t mind installing a little plug-in when it gives them content they actually want.

There also have been some speculations about Joost-enabled set-top boxes in the past, and some people have wondered in recent days whether this could be a way for Joost to leverage P2P. Honestly, I wouldn’t hold my breath for it. First of all, Joost never seriously developed a Linux implementation, which is a must for traditional set-top boxes. Secondly, a new generation of boxes plays Flash, and Joost has started to embrace these devices with its web API. The company’s content is already available on DivX Connected and Neuros boxes, and a Boxee plug-in also seems inevitable.

Speaking of APIs: One of the many unfulfilled promises of Joost’s client was its API, which in theory allowed pretty sophisticated interaction with the service’s users, as well as its content. The API had great potential, but hardly anyone developed for it because hardly anyone used Joost. It seems that the company now wants to get a larger audience with its web platform and then gradually re-introduce some of the client’s more advanced features.

Volpi promised that one of the upcoming releases of Joost’s service will feature the ability to tag and comment on specific points within a video. It also plans to open its web site to widgets from third-party developers some time in the future, but this “will require more development,” as Volpi told us.