[bittorrent] The value of trust and acceptance

Justin Cormack justin at street-vision.com
Sat Mar 26 07:12:08 EST 2005


On Fri, 2005-03-25 at 15:35 -0500, Simon Spero wrote:
>   The protocol and model have too many critical  flaws to teach us much 
> directly; it's the relatively large installed base that makes BT an 
> important and interesting system to study.

Could you say what you think these flaws are? I think a lot of people
have different ideas about the flaws (often depending on what they think
of the usage as being). The fact that the user base is so large suggests
to me that the flaws are not critical and can be fixed, but I would like
to hear your opinion on this.

> Secondly,  a sucessor content-distribution system that can offer a 
> gateway in  to these existing clients will find it a lot easier to 
> achieve criticial mass.   Such a gateway can be built by creating 
> pseudo-trackers that induce desired behaviour from the client by 
> manipulating the weaknesses in the BT  model for the client's benefit.   
> This approach meshes nicely with the perma-seed concept.

What client behaviours are you thinking of? Using the trackers to select
peers for particular clients? Can you elaborate on what you are thinking
about, it sounds interesting.

Justin





More information about the BitTorrent mailing list