[bittorrent] Protocol Deviation in Mainline

Elliott Mitchell ehem at m5p.com
Thu Mar 2 04:13:05 EST 2006

>From: Ludvig Strigeus <strigeus at gmail.com>
> It gives me the benefit that my requests will reach the recipient
> faster. Thus less request latency = higher download speed.

Bzzt! Wrong answer, thank you for playing.

If properly implemented, Bit Torrent isn't significantly bothered by
even fairly large latency (notice /properly/). No matter how great the
latency, you can send requests out for enough pieces to fill the BDP. The
only risk is that of how many you lose when the peer chokes you.

In fact 16KB pieces makes this worse, as you need to queue a much larger
number of requests.

> What's the big deal in using 16k requests anyway? We're just talking
> about like 30 bytes of overhead for 2 16k reqs instead of 1 32k req.

True, notice that hasn't been my main arguement. The issue is that you're
violating the protocol specification by disconnecting other clients that
request 32KB at a time.

(\___(\___(\______          --=> 8-) EHM <=--          ______/)___/)___/)
 \BS (    |         EHeM at gremlin.m5p.com PGP 8881EF59         |    )   /
  \_CS\   |  _____  -O #include <stddisclaimer.h> O-   _____  |   /  _/
    \___\_|_/82 04 A1 3C C7 B1 37 2A*E3 6E 84 DA 97 4C 40 E6\_|_/___/

More information about the BitTorrent mailing list