Cringley doesn't think Google is doing TV right

In a conviniently timed article, Cringley talks about how the server-client methodologies will not hold if internet distribution of video is to reach any ammount that would impress the media producers. Since I was just posting about Google TV, I thought I'd put up some related thoughts.

There are a lot of things to agree with and to disargee with. On the one hand, I can see that the current methods wont last if the userbase grows large enough to care about. On the other hand, I know those execs are afraid of the term "P2P" so lets keep away from it, or it just won't happen.

The majority of the content is pre-recorded, so we aren't talking about live streams; this is stuff that can be downloaded ahead of time and watched from your local cache, so that's in our favor. Also remember that current broadband can bring in a lot video faster than you can watch it, so the servers aren't even supporting the users for the entire duration of their entertainment time. This, at best, can double the number of viewers a data center can support.

So the current ways might live for a little longer than Cringley thinks, but in the long run he is probably right. Of course, this kind of thing won't be available easily and universally until its built in at a lower level. What we need is something like HTTP can operate very easily, be implemented quickly, and adopted by everyone. Most importantly, it needs to be transparent to everyone but the network admins and browser/player developers.

Grid Networks looks cool. When I get my new Windows box, I'll try out the demo and see what kind of content they have available.