Re: [tlug] Bittorrent Newbie

[Edward Middleton <edward@example.com>]

Stephen J. Turnbull wrote:
>     Scott> Oddly enough, though it's supposed to be faster, I've
>     Scott> always found it slower than standard downloads of ISOs.
>
>     Stephen> No, it's *supposed* to be slower,
>
>     Edward> It is supposed to be less bandwidth efficient, not slower.
>
> Stephen> No, it's supposed be more bandwidth efficient (ie, it uses unsaturated
> Stephen> bandwidth, which---except for packet-metered connections---is free),
> Stephen> and slower.  
That is not what bandwidth efficiency means.  For lack of a better
definition it is a relationship between the transfer speed to the
bandwidth used.  If you are talking about downloading a file the
upstream bandwidth is just wasted bandwidth (protocol overhead), which
is why the bittorrent protocol is less bandwidth efficient then FTP.
>     Edward> If you are using ftp to download a file your download
>     Edward> bandwidth is limited by the servers bandwidth.
>
> Which may or may not be the bottleneck; typically it is not for
> servers that offer FTP.
Bittorrent is solving a different problem to the FTP protocol. 
Bittorrent is designed to combat the slashdot effect that even the most
well provisioned FTP server can't, or don't wish to deal with (i.e. will
block with user or file size limits).
> It is true that bittorrent allows servers with piss-poor connectivity to serve larger files than they could by FTP, but by that very token you're unlikely to actually observe them in reality.
And CD and DVD piracy is unlikely to be observed in reality ;)
> and much of the time they are likely to be slower on the same hardware.
How about you say something concrete.  Who's hardware, are you including
the network?
>     Edward> If you are downloading with bittorrent the bandwidth is
>     Edward> limited by the number of seeding (I think that is the
>     Edward> correct terminology) nodes.
>
> Assuming (as you implicitly do) an infinitely large file or a random
> and growing distribution of active nodes on the Internet, and that you
> are "close" to the backbone, it's limited by the minimum of the sum of
> the bandwidths of the active nodes and the minimum of your upload and
> download bandwidths.  In other words, torrents are going to be slower
> on ADSL.
>   
In a perfect world were FTP server resources are unlimited and FTP
bandwidth limits don't exist.  In the real world it will depend on how
active the torrent is, and how loaded the FTP server is.
> NB, the special characteristic of a seed node is that it starts with
> the whole file but nevertheless it stays active until the torrent is
> deactivated.
>   
Well if this is the correct definition of a seed then I have use the
term incorrectly.  I was referring to nodes that don't block uploads.

Edward