[MLB-WIRELESS] Fwd: [mesh] WOAH!

[Paul van den Bergen]

alright, I have been biting my tongue so far, but I think it is time to 
weigh in (besides, most of the issues have been exposed sofar, which 
makes it easier to sound wise now!:-)

Seems to me that there are several issues here.  

We are building an Ad-hoc wireless peer to peer network (with short 
circuits). sounds like an ideal environment to do funky small world 
theory stuff.  This is an evolutionary proccess.  As long as it is fun, 
people will do it.  if it is not fun, people need a reward for bothering.

To ISP or not to ISP?  
I suspect something like this will happen.  Once we have something 
approaching a meshed network (*poink*)....
(several) someone(s) who is/are enthusiastic will connect their internet 
feed to the MW Intranet.  they will have some sort of access 
restrictions varying from eg. no requests but access to proxy/cache to 
unlimited (time, rate, datacap) access.

as the cost to these individuals goes up, they will get frustrated and 
drop it.

as users get frustrated with the increasing delay and poor performance, 
they will set up their own connection, request MW provide pay per 
data/time access (with costs passed on to members), which will no-doubt 
be argued about for a long while.

personally, I would love to have internet access augmentation of MW, but 
will still likely retain some other internet connection anyway (melbpc, 
for example).

I don't have an opinion one way or another, except to point out there 
are a lot of options between baring the public and letting anyone use it 
(eg.  paid up, signed up (authenticated) members get to make internet 
requests, everyone else is limited to proxy cache or distributed 
files...  there are other models too).  The level of access will change, 
depending on the average enthusiasm and wealth of the group.  maybe we 
will morph into an ISP, or not.  in a way we are already doing that - 
every one of us who has a public AP will be a mini ISP (well, intranet 
service provider anyway).

I cannot say that this debate is going anywhere.  We just don't seem to 
be able to see the shape of the future yet.

 I agree with the opinion that we are an attractive body for an existing 
ISP to extend their customer base.  but I doubt we could sell them our 
bandwidth - quite frankly, I don't believe we want to offer the QoS 
routinely expected from commercial operations.  And this is the key to 
understanding this.  We can certainly set up as an ISP to joe public if 
we want, but the level of quality control would necessarily be low. (or 
maybe not - it depends upon the individual whos AP you access).  as long 
as we are all aware that the QoS we provide is low (or non-existant) we 
can (presuming we don't run into cost over-runs - more on this later) 
easily provide ISP - like services - presuming regulartory issues are 
solved - not trivial, but in some ways moot.  Either they will be OK, or 
not (or eventually we will work around them - remember, the internet 
community sees censorship as damage and routes around it)


What other services (for want of a better phrase) can we offer? (i.e. 
why would someone want to join us?)

Data access and storage - aka ftp, BBS, etc.
this is a no brainer.  access to 100's of GB of data locally distributed 
for the cost of setting up the hardware....

Distributed network proxy to reduce internet connection download costs.
amortise the cost of internet access - ie. no internet access as such, 
just reduced load on your existing internet access.

Community and Fun
chat, gameLAN, other funky stuff (think small world theory, spam and 
antivirus, gnuttella and P2P).  I am personally looking forward to some 
interesting fun hardware hacking antennaes and messing with small world 
theory using the MW as a test bed (all your bases belong to me)

that last item is interesting.  this is a hacker community.  It is 
currently not a Joe Public network.  It may never get to that.  but I am 
more than happy to let the nuffers in and educate them (somewhat....  I 
am wondering if we will get overwhelmed later).  Also, experimenting 
with this beast will no doubt produce new ways of working, playing, etc. 
that I believe will have commercial potential (well, maybe not for us, 
but the advancement will benefit us all, maybe.  Well, it might be fun). 
Like what? you may ask.  dunno, else I'd be doing it already...


Public access models. (my prefered embodyments)

1)
free open internet access to anyone who wants it - member or not.

problem - who pays?
I don't think this will work.

2)
free open network (intranet) for anyone who wants it - member or not.

access to files (P2P/gnuttella style or BBS style), chat, forums, news, 
games, etc.

limited internet augmentation if distributed proxy caching can be made 
to work (well, in a regulatory sense. we know this can be done physically).

No internet access provided through MW

I believe this is the most likely outcome and worthwhile

3) as for 2, but with internet access provided for MW members where 
local POPs can be established.

This would be the ideal embodyment - we may have to do some fiddling 
with download limits with members, etc, but at least it'll be member we 
have to organise, not just anyone.


BTW, how many female members do we have?  why?  I can't quite get a 
handle on this.  Maybe I hang with a different crowd, but my geek 
friends are pretty much split 50:50 M:F.  On the other hand, it is 
mostly my male friends that play the one-up-man-game and tend to Herbert 
a lot more....

-- 
Dr Paul van den Bergen
SERC, RMIT University
paul at serc.rmit.edu.au
+613 9925 1624 phone
+613 9925 5699 fax
goofey: bulwynkl




To unsubscribe: send mail to majordomo at wireless.org.au
with "unsubscribe melbwireless" in the body of the message