[Syd-Wireless] RE: [MLB-WIRELESS] Is Melbourne Wireless dead?

[Dan Flett]

Hi Duane,

Duane wrote:

> Dan Flett wrote:
> 
> > The main attraction of a CWN is that it is Free.  Certainly the hardware
> > costs aren't free, but they are relatively cheap and are getting
> inevitably
> > cheaper.  And these are a one-time cost, whereas ISP fees go in forever.
> 
> This thinking is true to a point but ultimately flawed, for starters you
> need some very capable people able and willing to run a network, time
> here while it may be donated long term on a city wide project won't be
> free, things stop working for no reason, (based on the target
> demographic) things will need to be constantly upgraded to prevent the
> script kiddies from disrupting things too much, you will need constant
> monitoring and be reactive to issues that can and will occur like any
> other large area network.
>
> Sure you may get some enthusiastic people while they see it more of a
> benefit then a tax on their time, but long term it's going to end up
> costing in labour charges if you hope for longevity especially as things
> expand. As in most businesses you will find labour charges will far
> exceed any other expenses.

I see community networks as the network equivalent of Open Source software.
People contribute to Open Source because no-one is being exploited.  As soon
as someone sells software, then the person writing that software wants to
get paid.  If the software is being given away for free and is legally bound
to stay free (a la the GPL) then people feel driven to contribute in
extraordinary ways to the project.  The software becomes public domain and
contributing to it is contributing to the public good.

It is the same with a community network.  If no-one is making a buck from
it, then many talented people will derive satisfaction from contributing to
it, gratis.

Sure, not everyone gets satisfaction in this way, but the larger the
network, the more people benefit, and the more people will want to
contribute.  Contribution could be the day-to-day maintenance of the
network, or it could be developing smarter systems to automate the
day-to-day maintenance.
 
> Next, backbone hardware, connecting up mass amounts of nodes across a
> city requires some centralisation, you simply can't get away from it.

I totally agree.  I don't believe that flat "mesh" networks - which would
require hundreds of node "hops" to get from one side of a city to the other
- are scalable to a city size.  A hierarchical - a star or "central ring"
topography is needed.

> Not only will the backbone need to be stable it will require higher
> bandwidth availability to ensure that you don't end up with lower then
> modem speeds when everyone wants to be doing more then text based games
> and chat.

A community network would make no service-level guarantees at all.  It would
not claim to be fit for any purpose whatsoever, etc, etc.  This
higher-bandwidth backbone would be obtained through the donations of
philanthropic persons who believe in the principle of a free data network.
With the ACA regulations the way they are, the point needs to be made that
this donation is in no way payment for a service or any sort of commercial
arrangement being made by any organisation.  It must be a private donation -
or the equipment must remain the property of the donator.

Higher-order nodes would probably be something like Motorola Canopy,
KarlNet, WiMax gear or similar.  Stability would come when nodes are
plentiful enough that diverse routes can compensate for the failure of any
one particular node.

> While you may not be able to make money from it, I'm pretty sure you
> could get away with cost recovery, but once you start doing the maths it
> starts becoming a pretty expensive exercise and with the lower economies
> of scale from less users the costs soon start adding up to be similar if
> not more expensive then purchasing commercial services.

This is true if you think with a purely capitalist mindset.  Thinking with a
non-profit-organisation mindset you start with the principle that the
end-user pays for nothing other than their own node equipment.  This pays
for all the "last mile" costs.  Those end-users that think that this free
network is a great idea will want to contribute to the network by installing
a routing node - to further the spread of the network to other users.  This
pays for the second-order node costs.  That leaves the highest-order node
costs, which is left to the truly philanthropic node-owners who will donate
expensive high-capacity gear to get the whole city connected.  There are no
ongoing costs as all real-estate is donated, and any electricity and
maintenance costs are borne by the node-owner.

Certainly with this idea you can't make any sort of guarantee that any
particular area will be covered by any particular date.  You can't lay out a
specific business plan, you can only lay out general principles, do your bit
and watch what happens.  You trust that this non-profit network is such a
great idea that people will get on board and help out.  It is my aim, and of
many others, to prove that this principle works, and I will work hard to
make it so.

> The mobile carriers etc do charge like wounded bulls for a number of
> reasons, one of which is return on investment because the market place
> is changing so rapidly you end up needing to outlay more money for new
> technology before you've been able to repay for the last outlay,

And of course they need to keep their shareholders happy.  And shareholders
by and large are interested in returns, not necessarily contributing to the
public good - ethical investors notwithstanding.

There certainly is a place for telcos.  I don't propose that free community
networks will ever replace telcos, simply because people need a Service
Level Agreement when it comes to their data, and they will pay for it.  But
community networks provide an alternative that will keep the telcos in check
and stop them charging more than they should.  Telcos are reluctant to
innovate, community networks will experiment whenever and wherever they can.

> the
> other most obvious reason is of course lack of competition, but with dsl
> getting much cheaper along with wifi hardware and mobile phone makers
> and a couple of standards organisations tweaking their interest in
> wifi/voip and actually making a hand over protocol between wifi/voip and
> gsm and sending gsm packets over IP things in the next few years should
> get VERY interesting to say the least.

Indeed, which is why community networks are so important.  The will be able
to participate at the front line of this VOIP revolution, as they are 100%
compatible IP networks.  Community networks the world over are likely to
contribute much of the innovation in this area.

Cheers,

Dan

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