[MLB-WIRELESS] Re: mojo -was- structure

[Ben Anderson]

> While an interesting idea its flawed and is going to be to awkward to
> implement and police.

implementation difficulties yes.  Police?  That should be dealt with in the
design....  Protected by high encryption, and designed so that it's a stable
system, it shouldn't need people having any form of active control over the
structure of the system.

> and thats mainly because theres no default gateway or firewall for all
> traffic to pass
> through,

Which I think is a good thing.

> my personal link to x node wouldnt end up getting me credits in the mojo
> system,

Why not?  Any data you transfer on behalf of someone else should get you
mojo (and more if it's an 'in demand' resource).

> On top of that no ones running the same setup so what ever mojo client
> program
> that ends up having to be written is going to run into trouble, what
> with
> some people running ap's and others nix boxes with wireless nics...

Yes, this is an implementation detail.  It's a practical issue, which is why
I was looking at making a chunk of hardware with a fully enclosed system for
a couple of hundred dollars.  Better for the 'average joe' to make the stuff
work.  And people who want to hack a setup together with an old laptop can
still do that, just if they want to be part of the setup, they've got to be
willing to change the software.


> the other flaw is that why would members decide your traffic passing
> over there hardware
> is more important than theirs....

Because to become part of the system, they have to be part of the system.
If someone black-wholes themselves somehow, and ignores all packets except
their own, they're not going to earn mojo, and subsequently will be
restricted to using whatever bandwidth is available 'free'.  People with
mojo get 'almost guaranteed' service.  Unmojo'd people get whatever's left
over.

> this is a free system, so any qos you get is what ever is avalible at
> the time.

Again, what makes a system where low-latency is DoS'able by anyone doing
anything high bandwidth more free than a system that protects all types of
traffic?

> as far as realtime video / sound goes for video confrencing netmeeting
> runs pretty well
> on a 64kbs link and voice only is much lower again, and frankly 200kb/s
> is some kinnda fat ass
> application that atm shouldnt be running on this system unless you setup
> or work out a dedicated
> link across town.
> I dont feel running 30 fps for video confrencing is a good use of the
> system atm unless your planning
> on doing phone sex where high quality video is possibly a plus.

Emotionally charging statements... tut tut :)
The exact numbers, 64kbits, 1mbit, a gigabit, are not what I'm concerned
with.  Regardless what we put in, if it's "free" it'll get fully utilised if
it scales large enough.  And that means zero (none, nada, zip) low-latency
traffic gets through because somebody's leeching some large file (and
there's lots of large files around thesedays) 24/7 from across the other
side of the network.  And that leech is likely to be a leaf node, not
someone donating time, effort, money to set the system up better.  Why
should the few be able to destroy the network for everyone?  And if the
users donating the time & resources can't use their newly installed network
link to talk to the people they wanted to talk to, then why are they going
to spend _more_ money on it instead of turning public access off?

Ben.




--
To unsubscribe, send mail to minordomo at wireless.org.au with a subject of 'unsubscribe melbwireless'  
Archive at: http://www.wireless.org.au/cgi-bin/minorweb.pl?A=LIST&L=melbwireless
IRC at: au.austnet.org #melb-wireless