BitTorrent + RSS = Broadcatching

Well, if you tried distributing any kind of big file on the Web, you've learned the pain of bandwidth expense, and the difficult trade off between filesize and quality. As Grumet points out:

The same motivator for enclosures---to have the downloading happen overnight when you're (hopefully) not sitting at the computer---applies. It's just the details that differ.

I can use "Nothing So Strange" downloads as a practical example. Between transaction charges (15%) and high-quality hosting bandwidth ($7/GB transfer) providing even a medium resolution (ie, as good as television, not as good as DVD) for a film is 466MB in size. That means we bring in $4.25 net on each transaction, and the bandwidth costs us $3.26. That means we "made" a whopping $0.99 cents if they have a successful download the first time, and we lose up to $2.27 if they have to download a second time.

If we wanted to make a downloadable version of the DVD available, we could certainly do it as easily as the above video file ... and with over a million DVD-R drives out there, it might even be a viable method of distributing. But now we're talking a file more like a gig and a half in size (or $10.50 in bandwidth costs.) That's twice as much as the cost of a "just in time" duplicated DVD, and ten times as much as our per-unit cost on the "Nothing So Strange" DVD run that we retail for $14.99.

These margins and the edges of cost and value are a hamper on the real blossoming of video distribution on the Web, and can only be aggregated so far out of the way. P2P swarming technology is the only current viable route to break that stalemate by spreading at least part of the costs away from your own bandwidth pipe, but under a system like BitTorrent that's only really useful if there are a lot of people with fully download copies to swarm from (so you have a classic tipping point model of efficiency.) Promotion preceeds adoption preceeds efficiency.

The brilliance of an RSS approach, though, is that it builds in at least two important features that BitTorrent alone doesn't address. First, it provides a method of propogation through editorial filters -- a successful editor picking new BitTorrent works could help create an instant rush to the tipping point, in the process decreasing the cost of bandwidth on each copy. Second, it turns BitTorrent into a subscription system, one where your system automatically collects new content of a large size overnight (for example.)

Of course, there's no gateway -- the idea of charging for that file is still an alien concept. So while it's certainly not ready for primetime, it's an interesting area for video distributors to keep their eyes on -- not so much because BitTorrent itself is an emerging system, but because the idea of download swarming coupled with syndication represents a powerful synergy for the larges of big media objects.