Comments on: New BitTorrent could save ISPs billions — UPDATED

By: Nick G

Nick G — Wed, 04 Nov 2009 13:07:09 +0000

…or at least, it *should* work like that. In practice, it seems it’s poorly implemented and has no significant effect (according to the link in Dave’s comment above).

By: Nick G

Nick G — Wed, 04 Nov 2009 12:53:34 +0000

Fibre1 – please don’t comment if you don’t understand the protocols or traffic shaping. It does not reduce bandwidth – it reduces the file transfer rate so that fewer packet collisions occur and there is less congestion. Therefore there are fewer retries and overall, the transfer rate is faster for a connection of the same bandwidth. It works the same way that variable speed limits on motorways improve traffic flow when there is bad congestion – if you eleminate the few idiots trying to drive at 60 in traffic by lane-weaving and causing the snake or ripple effect then the traffic flows better when limited by law/cameras to 40.

By: Dave

Dave — Tue, 03 Nov 2009 20:20:33 +0000

There’s no data to go on as yet – only BitTorrent have the figures from their beta testing.

If you’re interested, there’s a more technical analysis here: http://www.digitalsociety.org/2009/11/analysis-of-bittorrent-utp-congestion-avoidance/

By: Fibre1

Fibre1 — Tue, 03 Nov 2009 09:21:38 +0000

“It’s thought user download speeds won’t be slowed down in the process ” – sorry, this is nonsense. If it reduces the bandwidth used it has to reduce the download speed, period.

Just another pirate tool looking to evade the restrictions.