[MLB-WIRELESS] I nominate myself for President

[Dan Flett]

Hi all,

I am nominating myself for the position of President on the Melbourne
Wireless Committee.  I am seeking someone to second my nomination.

I will be unable to attend the AGM on 9th February due to work commitments.
And it is likely that throughout the rest of this year, I will find it
difficult to attend a lot of the General Meetings, again due to work
commitments.  Unfortunately it seems to be the case that the people who are
most motivated to serve on the Committee are the people with the least
amount of spare time on their hands.  Nonetheless, I believe that as
President I will be in a position to re-invigorate Melbourne Wireless.

I want to give a direction to Melbourne Wireless

I want to make Melbourne Wireless more accessible, cheaper and easier to
participate in.

I plan to emphasise and promote a type of wireless network which to date has
been either unknown or unpopular inside Melbourne Wireless: Mesh networking.
But I do not intend this new approach to be at the expense of our
traditional approach: Infrastructure networking.

There are two main styles of wireless network - Infrastructure and Adhoc,
also known as Mesh.  

Infrastructure involves a network of point-to-point links using directional
antennas, and, ideally, a radio dedicated to each individual link.

Advantages:
Long distance links
High Throughput on each link if link is maintained and tweaked

Disadvantages:
Expensive - needs separate radios for each link
Time Consuming - each link needs to be planned, equipment bought and set up
and antennas pointed and maintained, needs high towers to get best distance
Ugly - some people don't like the idea of a big tower with lots of big
antennas on their roof
Steep Learning Curve - requires a decent level of knowledge about RF and
network principles

Mesh networks use omnidirectional antennas, with each radio making a link to
every other radio within range.

Advantages:
Cheap - a single radio can make multiple links
Easy - hardware and software can be designed to be virtually
configuration-free
Not So Ugly - Mesh Radios can be very small and unobtrusive

Disadvantages:
Limited Range - without amplification the cheaper mesh radios, if mounted on
a roof, have a range of about 500M - 1KM
Limited Throughput - Mesh topology cannot deal very well with many users
concentrating traffic on one node - e.g. an Internet Gateway.  Traffic needs
to be distributed, with care taken to avoid large bottlenecks

Typically, proponents of Infrastructure networks criticise Mesh networkers
and their methods as being unrealistic - the short range and lack of
scalability they say kill the viability of such networks.  And yet, all over
Europe community-based non-profit mesh networks springing up in the cities.
Not only do they just survive, they thrive.  The Freifunk network in Berlin
at last count had well over 400 nodes and seems to be growing rapidly.  Mesh
networks are also successful in the US, both commercially and in community
groups.  In fact, some community efforts have been successful enough that
those involved have chosen to commercialise their software and develop
specialised low-cost mesh hardware (the Roofnet/Meraki projects).

The saying Different Strokes for Different Folks is true in Community
networking; Infrastructure and Mesh networks appeal to different kinds of
people.  The challenges are different.  I believe that emphasising Mesh can
serve as a good entry-level for new members to get their feet wet.  Most
will probably not expand on their single-radio mesh node, but many may go on
to try a more Infrastructure approach.

Mesh and Infrastructure need each other - Infrastructure for its own sake
has limited appeal.  Mesh is not very scalable without long-distance links
to link small scattered meshes together, and to cut the number of hops
within large, connected meshes.  Combining the approaches gives the best
results.  By encouraging both approaches under the banner of Melbourne
Wireless, we allow our members to choose their own path and contribute in
the way they are most comfortable.

Playing with multiple WiFi radios with directional antennas and tall masts
is fun, but requires a steep learning curve or a high level of assumed
knowledge about networking and radio principles.  It is also expensive and
time-consuming.  For many this is a hobby unto itself, for many others it is
too much of a hurdle just to hook up to a network.   

Why should someone put up a wireless node?  The traditional answer is "to
connect to the network".  With Mesh networking, in the early stages at
least, this answer would seem to be a more than a little bit hopeful;
Standard roof-mounted Mesh radios have a range of between 500m to 1Km at the
most - so with no other mesh radios within this radius, a single-radio mesh
node-owner cannot connect to the network.  So why put up a mesh node?  

We need more reasons than just network connectivity.  Can a lone mesh node
be a useful community resource?  I believe it can.  I believe a lone mesh
node can act as a Community Hotspot.  It could offer Internet access through
the MelbWireless Hotspot package - but it doesn't have to.  The node's
captive portal webpage could do other things aside from authenticating
users' Internet access.  It could host community forums and a community wiki
- a local bulletin board.  Once people using the forums get to know each
other, other activities could be organised and services hosted on the node.
Even embedded nodes running on limited hardware can do quite a lot with just
8Mb (or less) of Flash RAM to store the software.  The point is we should
aim to make Melbourne Wireless nodes useful and easily accessible to the
wider community, not just to other Melbourne Wireless members.

None of these suggestions are new.  We've seen all sorts of interesting
services offered via the nodes of Melbourne Wireless members.  But to date
these efforts have all been of the "roll your own" variety.  Some are very
successful, many shine briefly then seem to die out, and a lot never get off
the ground (often literally).  The amount of effort to develop a fully
featured (and integrated) set of services is often underestimated.  Many of
us who put in the hours, days and weeks building a node have a good time
learning how to do it, but wonder "what next?" once the node is built.  Once
the experience has been experienced, there's not a lot of incentive to go
through the process again.  I would never discourage anyone from attempting
the "roll your own node" project - it is a great learning experience.  But I
feel with the weight of these experiences behind me, that Melbourne Wireless
should offer a quick-and-easy node-building option to its members.

That quick-and-easy option is best offered in the form of a standard
firmware image for wireless routers that incorporates a set of services that
achieves the aim of creating a Community Hotpot.  With a standardised
firmware, tailored to suit Melbourne Wireless and distributed through our
website, we can mass-produce nodes according to what we think a "standard"
node should do.  We can benefit from each other's experiences and knowledge,
and put all our best ideas into the firmware.  We needn't re-invent the
wheel each time a new node is put up.

I believe there are people willing to share their Internet for free if it
can be made easy for them to do so.  People will be more likely to share
their Internet freely if they can interact with, and get to know the members
of their community who use the Internet feed they are providing.  There is
little incentive to give one's Internet bandwidth away to anonymous
leechers.  There will always be leechers, of course, but they can be managed
in software.  I would hope that at least some of the Internet users would
find the concept of the Community Hotspot appealing and choose to
participate in it.  In any case, we should aim to spread the SSID
"melbournewireless.org.au" across the suburbs.  And I want users to greeted
with a friendly, informative captive portal webpage when they associate.
This will demonstrate a functional, vibrant network which will do wonders
for our public image, and will help boost our membership ranks.

I prefer "right-now" solutions.  I want to get working solutions out to our
members as soon as possible.  If a suitable or almost-suitable solution
exists already, anywhere in the world, we should do the minimum amount of
work required to adapt it, and then use it.  

Development of new whizz-bang technologies should occur only after we have
released something that works now.  Without a working system in place,
development seems lose momentum and stall.  More and more time seems to be
spent contemplating how to make the system as good as it can be, without
much actual progress taking place.  I myself have been especially guilty of
this.  I now realise more than ever that having a working system in place,
even if it has flaws, is the best way to spur the development of even newer,
better systems.  

We are now well-placed to put this plan into action.  Wireless routers that
run Linux are continually dropping in price, are getting smaller, and are
being sold with more features that lend themselves to this sort of Community
mesh.  There are even a few commercial ventures hoping to profit from
exactly this style of networking.  But I believe it can be done for free.  

That's my plan for Melbourne Wireless.  I know that not everyone agrees with
it. I know that it probably doesn't appeal to a sizable percentage of the
current membership because it promotes a type of networking that is quite
different to that which most of us joined to participate in.  But I believe
that Mesh networking will regenerate interest in our existing Infrastructure
network.  It will provide a much-needed shot in the arm for Melbourne
Wireless.

Even if you don't like my plan, remember: A bad plan is better than no plan
at all.

On other topics:

It has been suggested that Melbourne Wireless raise its membership fee to
help pay for equipment to build a backbone network.  In principle I agree.
But I believe we can't raise our membership fees until we have proven that
joining Melbourne Wireless is worth paying extra for - we must first
re-invigorate Melbourne Wireless and rekindle a more mainstream interest.
Once we have a track record for getting things done, we will be justified in
asking for a higher fee.  If we raise our fee now, we risk losing our
members at an even greater rate than at present.

In the near future I do believe we need to look at other ways to raise
funds.  Any idea is a good idea, no matter how old-fashioned.  I like the
idea of raffling off donated wireless gear at General Meetings - but I'm
sure the membership can come up with something else!

I would like to take the opportunity to answer questions and debate any
topics that any member wishes to put to me in the coming week.  Remember you
need to be a paid-up member to vote!

Vote Dan!

Cheers,

Dan Flett