Norwegian Broadcaster Evaluates BitTorrent Distribution Costs

FrostPaw writes "An experiment was conducted recently by Norwegian broadcasting company NRK involving the release of the series 'Nordkalotten 365' (a wildlife program) in a DRM free format using BitTorrent. One of the broadcasters has posted the approximate figures for the overall distribution costs, and discussed his reasons for doing so. Their estimated cost for using Amazon S3 to offer the files through HTTP/FTP/etc. come to approximately 41,000 NOK (about $8,000 US). However, when using the Amazon servers as the originating seed and utilizing BitTorrent, their total cost for distribution of the entire project, thanks to generous seeds, would amount to approximately 1,700 NOK. The post with the original figures is available only in Norwegian.

Comments

This Just In:

[rsmith-mac]

Making other people do your work for free makes your own costs cheaper. Film at 11.

In other words, why is this news? It's something that has been obvious about BitTorrent since day 1: if you can get/make your users use their own upload bandwidth, you won't need as much of your own, and in a cost model that means your costs are lower. Did this really require a study?

Re:This Just In:

[yakumo.unr]

It's news because a lot of marketers need it spelled out for them, with big juicy numbers with currency symbols attached, once they start to really realize the financial positives of using the most efficient distribution systems, they might stop trying to shut down just that, a highly efficient distribution system. it's not the personification of piracy.

No such thing as a free lunch

[drhamad]

At some level this is redundant, but I'm going to state it in a slightly different way.

Of course distributing via BitTorrent is cheaper for the originator, nobody could possibly argue this. But I'd like to see a study on the TOTAL cost to society. In other words, yes it's cheaper for the originator, but there is no such thing as a free lunch. SOMEBODY is paying for all that bandwidth/etc. If you have bandwidth limits, perhaps you are paying for them to distribute their file. If you don't (as we in the US do not) then the telecommunications company is paying. Bandwidth does not materialize out of thin air. SOMEBODY pays. Further, BitTorrent is not exactly efficient. It uses a lot more requests/connections/etc to download or distribute via BT than it does via HTTP/FTP/etc.

The offsetting factor may be the more distributed load over the system, since there's no central point, really. I'm not sure how much this really helps though.

I guess my point is, the total cost to society of BitTorrent use may very well be higher than that for distributing by older methods.

A 100% share ratio requirement is unrealistic

[tepples]

And you can use a tracker that will ban users if they don't seed to atleast 1:1 It is mathematically impossible for everybody to seed more than 1:1. That would require the sum of uploads to be greater than the sum of downloads, when they're supposed to be equal by definition. Besides, for an older file that has 20 seeds and 0 downloaders, how can one seed to 1:1 without keeping the computer turned on and connected to the Internet for weeks at a time, praying that another downloader might show up?

Multicast?

[gjh]

What I'd really like to see is figures for the broadcaster and the hidden costs to the ISP for each of....

- Unicast
- Bittorrent
- Multicast

Multicast is so obviously the best solution all round for the, what, at least 50% of a national TV station's audience that watch predictable and consistent shows week after week. It would be pretty trivial for PCs to grab a multicast overnight.

By the way, the BBC really tried to do this right, but ISPs were too stupid to see that it was in their best interests to cooperate. This is my reading of the evidence - I accept corrections.