Re: Bittorrent (rate limiting vs. prioritization)

[Martin Fick <mogulguy@xxxxxxxxx>]

--- On Wed, 2/18/09, slush <slush@xxxxxxxx> wrote:

...rate limiting ... prioritization...

> I think it is very similar view to the same problem and it
> depends on ease of possible implementation. 

I don't quite agree, they really are two very different
solutions to address two different problems.  Prioritization
addresses a threat to the tor network itself, whereas rate 
limiting addresses using tor to abuse others.  So, it is not 
just an ease of implementation issue, but two very different 
agendas.

...
> Since there are very good legitimate reasons to want to
> > email anonymously, a very low bandwidth rate for email
> > might still make it usable for emailing anonymously
> > but not for SPAM.
>
> Im not sure it will work. 

Well, of course, neither am I. ;)

> When you tight up port 25, there will be the
> same(similar) amount of spam, 

I disagree, as soon if you keep the total tor network
bandwidth allocated to a very small number on port25 
(to exagerate, say 1 K per hour), there is very little 
reason to believe that spammers would use tor at all.  It 
would be completely useless to them.  However, it may 
still be useful to a whistle blower who really needs to 
stay anonymous since he may eventually get his message 
through and he now has a larger choice of exit nodes
to choose from to protect his anonymity.

> ...but because of speed restriction, there will
> not be enough connectivity for regular users :-). 

You may or may not have enough bandwidth left for normal
users, but there is a chance you may actually have more
since you will have hopefully eliminated the spammers
use of the tor network (are there actually any?).  The 
objective is to make it possible to keep the same amount 
of bandwidth for port 25 as there currently is available, 
but to better distribute that bandwidth across exit nodes 
to improve anonymity without incurring a greater spam hit.

> With priorities, you will be able to handle big amount 
> of request, when your exit will have free capacity.

Exactly what I don't want for port 25 since it will
only encourage spam usage of the network.


> But in fact, I welcome any method of QoS - both port speed
> limiting or port priorities. The best is both, of course ;).

Agreed!!! ;)


-Martin