The YouTube Question

With all of the press and talk these days about YouTube I'm sure you guys have checked it out by now. We love it because we believe it's the first step to giving our users free video content without having to make huge deals with media companies. It also fits into our philosophy of giving control to the users. You should be able to control your content and distribution, not just the big media giants. That's a scary thought for them as things continue to grow for YouTube.

The technical side of why I'm putting up a post on YouTube has to do with bandwidth costs. Right now they're burning about 1.3 Million Dollars a month on Bandwidth alone. Thats nuts. And even though I'm sure YouTube won't go silent because of all the eyeballs that are constantly on it, I am concerned with how a public CDN like this can remain viable with such huge costs? I understand that it's only a matter of time before we start seeing ads embedded in the videos like the ones we see on Cable and TV - you know the ones that run in the lower corners of your screen and annoy the crap out of you? Anyway, with ad revenue anything is possible. There's just so much money in the industry that it makes even insane bandwidth costs like YouTube's seem workable.

Contextual mapping across our device and system areas means we're going to include uploading videos to your deeda account a possibility, but is there a more efficient way to deliver it that reduces costs - that is the real YouTube question.

What I started to think about was P2P networks - offload a thousand times and then have those users become your content delivery vehicles rather than your servers. Torrents and P2P technologies also need a beauty and image makeover. What's ugly about them is that they're only accessible and intuitive to people with a somewhat technical background. YouTube proves that once you can streamline everything so it's one-click simple, people will come in droves. So why can't we hide everything from plain sight and make a P2P system that works seamlessly in the background of a YouTube-style website?

Cache Redirecting is what I was calling it when I first started working on it -- but the idea was to simply take packets out of a browsers cache, tag and then dump them back out to the web where the next user could pick them up to watch the same video. Then I came across these guys:

Solid State Networks

What's so cool is that these guys have already done all the dirty work for us. Larry Marchman was kind enough to show us around their technology, and explained how we can use their solutions on the delivery side of our video streaming, reduce bandwidth costs by 50%, and hopefully give our users a better online video viewing experience without buffering and jerking during high site traffic.

The only question to a hidden P2P system is this:

Would you mind clicking and doing a one-time-only install of a plug-in so you could benefit from this technology? Basically, all deeda.com users would have to install a one-time plugin before being able to watch videos -- this part doesn't seem ideal to me, but its still worth keeping in mind as a consideration.

Tell us your thoughts!