[wsfii-discuss] fon/meraki/community meshes differentiated .. (was: Re: Fwd: wireless connectivity.. Is802.11n the end of ethernet?)

[Julian Priest]

many thanks to all on this thread. 

Here's an addition. 'scuse the length of the post it's been a
while since I wrote on telco subjects and revisits many old themes.

On Tue, Oct 02, 2007 at 12:27:49PM +0100, vortex at free2air.net wrote:
>
> FON is a well funded company offering a pyramid scheme of bandwidth and access 
> control services utilising an AP model for maximum connectivity to potential 
> income streams. Some FON firmware versions have been cracked (thanks to 
> sven-ola and others) to allow freifunk mesh networks to openly pass thru and 
> ride invisibily along side an otherwise operable FON AP [1]. They also claim 
> to be the "world's largest wifi "community" [2] (we'll come to that "c" word 
> in a moment).

* Tying the knot

As if on queue, this announcement from Fon and British Telecom.

http://www.guardian.co.uk/technology/2007/oct/05/internet

"BT has teamed up with Spanish technology company Fon, backed by
Google, to try to persuade its more than 3 million British internet
users to open up part of their home wireless broadband networks so
that other people can use them."

http://blog.fon.com/en/archive/general/btfon041007.html

"From the very beginning, all of you, Foneros, believed in the concept
of sharing and in peoples ability to build something important that would
benefit everyone. BT is one of the most important telcos and ISPs in
the world, so with BT FON those beliefs have proved to be
well-founded!"

doh! at the risk of falling for this obvious troll :)

The publicity presents FON as a freenetwork. Indeed the rhetoric,
wireless hardware, and even software distribution (openwrt) are the
same.

* Free != Free

As outlined on this thread main differences between FON and a
freenetwork are:

1. Topology - FON is an extension of the DSL network and doesn't
   provide alternative routes apart from through the backhaul 
   provider. (This in contrast to an open local network like guifi 
   or a mesh like freifunk)

2. Ownership - who owns the box, both legally and in the sense of
   having root access.

3. Billing - these two create the possibility of a billing
   infrastructure at the node - the star topology makes the node a
   control point for it's clients, and the BT/FON owned captive portal
   uses this control as the basis of a billing infrastructure.

In the case of la fonera;

	free!=free 

or more clearly; 

	gratis!=libre

The 'consumer' gets the box for no payment but it runs someone elses
software, and 'consumers' are legally discouraged from changing it by the
T&C's even if it is reflashable.

Your freedom to config is limited *by* the gift of the device - the
capital cost of the device is recouped later through monetisation of
the portal that it creates.

In that sense it's a bit like a free web service - you can use the
service (built on free software platforms even) for no payment, but
you can't tinker with the algorithms and the economic benefit accrues
to the service.

* History

This announcement has some history. Going back to the UK in 2000,
wireless groups such as free2air.org and consume.net proposed wireless
networks as a way of building a self provided network infrastructure
whose ownership was in the hands of it's users.

Partly this was in response to BT's foot dragging over the
introduction of DSL based broadband. Partly it was an attempt to move
away from the monolithic vertically integrated near monopoly that
characterised network ownership then.

We figured that the Internet was too valuable a resource to be owned
by a small number of entities whoever they were(and still do). Rather
than following a competitive model between a few large providers (as
operates in mobile telephony), we proposed a distributed ownership
model based on collaboration or federation between many small or very
small networks. This ownership model reflects the decentralised models
made possible by the Internet protocols.

BT - wedded to a service provider model and trickle-feed bandwith
selling - perceived wireless as a threat and made several intervetions
into the space to challenge attempts to embed new network models over
the years. 

Normally BT have eventually adapted to the perceived threat to embrace
part of it as a business opportunity.

In 2002 BT lobbied against self provision, then switched to
lobby for commercial access to license exempt spectrum, then rolled
out a hotspot service (open zone).

As wireless developed, clusters of users would use wireless to
connection share. This was originally seen as a threat by BT however
they soon recognised it as a way of informally agregating users to
reach price points below those which BT could afford to market
to. (sub 5 GBP per month)

At the outset of the UK municipal wireless movement around the Access
to Broadband Campaign which cristicised lack of rural broadband, BT
responded by magically DSL enabling previously 'un-economic' exchanges
just as rural wireless was beginning to get some traction and
political backing in the UK. Then went on to bid for municipal
wireless projects themselves.

In 2004 we saw the acknowledgement of the 'default freenetwork' formed
by people leaving the default 'open' settings on their access points
providing essentially gratis hotspots. This was initially seen by BT
as a threat and was countered by a wireless security camapign but is
now being re-examined.

It looks like the rationale for this FON tie up is to give away
devices with controlled default settings and limited free access thus
turnning what they saw as a profitless network segment into an
opportunity for their hotspot business to grow. Now that BT have no
mobile arm, it is also a way of extending their Fusion home wifi/gsm
phone service and reducing reliance on third party GSM.

* Beyond Telecom

In a weird development in 2005 BT approached me to propose a research
project through BT research to look into ways of engaging the
community in operation of the local loop. (I declined)

It has long been recognised that British Telecom had a problem with
it's DSL technology based on a copper local loop. In the 80's to avoid
the goverment asking them to open their metropolitan ducting to
competing fibre networks, BT apparently completely filled the ducts
with copper thus physically preventing others from using the ducts.

Now having bet heavily on DSL, BT are tied to an ageing copper local
loop which requires an expensive work force of about 9000 just to
permanently maintain it - and maintenance is not really hitting their
targets.

"Unfortunately, the underlying failure rate for mainstream repair
performance (i.e. SMPF/MPF/WLR) continues at an unacceptably high
level (15-25%) & remains substantially short of agreed targets."
						Sept 2007

http://www.offta.org.uk/updates/otaupdate20071005.htm

At the same time broadband access has become a commodity market so the
margins are slim - too slim apparently to maintain the existing copper
properly or to roll out the more reliable fiber to the home.

The profit centres in the telecoms business have moved away from
consumer network provision, to wholesale backhaul, services at the
edge of the network (like google) and value added services.

BT are trying to make the move and increasingly focus on high value
services like Network Security and Software Development eg. the recent
NHS software contract.

At the time of their research proposal to me it seemed to me that BT
were looking for ways to dump the responsibility of the local loop and
their universal service obligation onto the community via wireless at
the point at which it had become an unprofitable burden to them.

I was left wondereing if BT would follow BP and rebrand Beyond Telecom
for financial reasons!

* Opposites Attract

It's ironic to see BT finally countering the perceived 'threat' of
freenetworks by absorbing the freenetwork movement's rhetoric and tools
(if not the substance of network ownership and topology) via the FON
tie up.

The press touts the FON/BT tie up as a triumph for open networks - and
perhaps it will extend access somewhat. However it is perhaps
better characterised as an attempt to increase the reach of BT's
vertically integrated network to compete with GSM in the mobile/voip
space via their Fusion product and to regain the outer edge of the
network as a billing infrastructure.

That said wireless freenets and BT/FON have become very close in
message and implementation. I'm not sure that if I was BT/FON I'd be
relying on the inviolability of an already compromised platform
created mainly by freenetwork developers to carry the weight of my
billing infrastructure!

On the other hand if BT/FON do manage to roll out as planned, up to 3
million devices will be well placed to implement an open wireless
mesh local loop network and only a few entries in /etc and a couple of
daemons away from doing so.

If history repeats itself BT/FON will be seeing that as a threat about
now, before adopting it as a key strategy before too long.

* C for Collaboration

For community networkers though, focusing on the other great, proven
network models like guifi and freifunk is going to be much more
productive in the long run than fiddling with foneras.

With the low cost of devices, open platforms like openwrt and the new
open hardware in development, there are plenty of options for adding
to rapidly growing community networks and adding to the common pool of
network, knowledge and tools at the same time.

Networks are all about co-existence and collaboration - succesful
infrastructures bind us together in subtle ways irrespective of
culture and economic goals.

For me the most interesting developments at present are in open local
access networks like guifi.net. Guifi have an impressive region wide
peer produced network, which features impressive volunteer
action, and even small businesses that aid people in maintaining the
network. It has bandwidth donated to it for social inclusion reasons
by local municipalities, companies and indivdiduals.

Ten years on it's great to see the model propossed as theory by
early freenetworks, borne out in practice in catalunya - 4700 nodes
growing by 50-100 a week.

Intriguingly at Guifi there is also the beginnings of telco uptake of
the freenetwork. With a free local access network companies are able
to offer chargeable services across it such as internet access,
guarenteed backhaul, voip or vpn. An open local network creates a
space for many players, both social and businesses, to offer services in
a way that a vertically integrated market like BT/FON never can.

* Loopback

Over the years there has been an ongoing unspoken dialogue between
freenetworks and BT in the UK with innovation in freenetworks followed
by adaptation and adoption by BT.

BT/FON has now shown itself to be just another the extension of the
old provider model. How much stronger the position would be for all
parties, if it was the *substance* of freenetworking that was being
absorbed rather than just the rhetoric appropriated once more as
marketing.

If history repeats itself it may not be long before BT sees FON as a
way out of their copper quagmire. If they see open local access
networks as an opportunity and not a threat it could be that they
themselves make those changes in /etc and install mesh daemons on
foneras!

Until telcos have the courage to see open access network models as an
oppotunity rather than as a threat, communites are better off
understanding and higlighting the differences and finding their own
way towards sustainable network solutions. 

cheers

/julian

[*] this post may include non-binding opinion, hearsay, forward
    looking statements, backward looking statements, delusional
    optimism, hopeless misrepresentation etc. your mileage may vary.