by Sasha Magee‚
Jan. 29‚ 2007
Much has been made of the proposed agreement that San Francisco has
negotiated with Earthlink that would allow the company to build a
citywide Wi-Fi network. But the coverage in the media has been
incomplete, and often incorrect.
San Francisco would benefit from a citywide network because it could
provide broadband internet access to people and communities who have
been left out of the rapidly snowballing internet revolution. Many San
Franciscans have no home broadband access for one or more of these
reasons: (a) no one provides it to their neighborhood, (b) a broadband
connection is too expensive, (c) they have no computer, or don’t know
how to use a computer, (d) the things they would get from the internet
aren’t available in their language, or at their reading level, or (e)
that they’ve never seen the need.
To address these barriers, the city created a broad
strategy called “Tech Connect” by the Mayor, which takes a four-fold
approach:
1. Create a citywide broadband network.
2. Provide low-cost computers to low-income and historically disadvantaged communities.
3. Provide training and support to people who need it.
4. Ensure that City content is available in a variety of languages and
at a reading level accessible by the bulk of San Franciscans.
An advisory Task Force, the TechConnect Task Force (of which I am a
part) has created a strategy to fulfill items #2-4. Although short on
implementation details, this report can be found here:
The key to this strategy, though, and what’s garnered the most press,
is the network. The City, after a series of interim stages, has
negotiated a tentative agreement with EarthLink to build a mesh-based,
wi-fi network to cover San Francisco. For approximately $22/month, San
Franciscans could get connection speeds of 1 Megabit per second
(slightly faster than DSL, substantially slower than cable modem). For
free, Earthlink (through a deal with Google) would provide 300 kilobits
per second connections, which is significantly faster than dial-up but
much slower than either DSL or cable. Earthlink would build the network
at no cost to the city, in exchange for up to four 4-year contract
terms.
At first reading, the proposed agreement sounds great. Free wireless!
At no cost to the city! It’s like free ice cream, only with no
calories! But there are some serious problems with the proposed
agreement, some of which I’ll detail below.
Coverage requirements are very vague. There are several
neighborhoods where it’s currently impossible to get broadband internet
access, because the companies involved don’t see the potential for
profit in those areas. Any effort to create a truly citywide network
must ensure that every neighborhood gets covered, and that the coverage
is equal everywhere. In the proposed agreement speaks about “whether
network coverage, quality of service, and performance are provided on a
non-discriminatory basis as agreed to in the Final Network Design
Documents.” I am assured that the “non-discriminatory basis” refers to
ensuring every neighborhood gets equivalent coverage, but it seems a
thin peg on which to hang the city’s whole strategy to provide internet
to everyone.
Many people will not get the network at all. The wireless
network, as planned, is likely to only reach the outside rooms of
buildings (if that), and only up to the second floor. This means that
anyone living in an interior room (one without windows), or wanting to
use the network in an interior room, will need some sort of wireless
repeater, which currently ranges from $50-$150. Earthlink will supply
premium subscribers with one of these repeaters (called CPE’s), but
people looking to use the free service will be responsible for their
own.
The network’s pretty slow. Without testing it’s impossible to
know the true performance of the network, but it seems likely that the
performance in any area will degrade fairly quickly if multiple users
are putting stress on it. If you’re trying to use the network to call
your family in Guatemala and your call keeps getting dropped because
your neighbor is watching a School Board meeting, the usefulness of the
network declines pretty rapidly. These are some of the uses common
today. It’s almost certain that by the time the network is built, there
will be many even more bandwidth-hungry applications that will be
impossible on the city’s free network. Even the paid level of service
is only as fast as cities like Mountain View are getting for free.
We’ll be stuck with it. Although the agreement is not officially
exclusive, the wireless radios will effectively saturate the spectrum
used by wi-fi. That means it’ll be nearly impossible for another
competitor to come around and provide cheaper (or faster) wi-fi
service. It’ll also mean that it’ll be difficult for the residents of
an apartment building or a company to run their own network.
The money we’ll get from it is pretty trivial. The city estimates that
Earthlink will pay approximately $300,000/year in fees to the city,
which are intended (but not guaranteed) to go toward digital inclusion
programs. This is the equivalent to roughly 1 minute per day of time on
the city’s parking meters. If we are serious about providing every San
Franciscan the opportunity to access the internet, we should allocate
the resources it will take -- not just what we can pry loose from a
company we’re giving effectively exclusive rights to our airwaves.
The privacy provisions of the deal are shot through with holes. Esme
Vos of Muniwireless.com, one of the world’s leading municipal wi-fi
analysts, describes the network use policies of the agreement as
“creepy”, and the privacy policies as “even creepier.” She sees the
agreement as potentially allowing “anyone who advertises on the
EarthLink network or anyone indeed who provides products and services
to any Fee Subscriber on the EarthLink network… to use your personal
info for other purposes, just not to send you marketing pitches.”
That means if you pay for an account Earthlink, Google, and potentially
anyone who advertises on the network can track your every move. Not
just every site you go to on the internet (which is bad enough) but
everywhere you go in San Francisco, unless you turn your wireless
device off.
The deal will likely last for a long time. The agreement specifies at
least five and a half years, which could potentially extend to 17 and a
half, if the agreement is repeatedly renewed. There are very few things
that Earthlink could do to not fulfill the contract. As long as it pays
its fees and has at least one wireless node providing free service, it
will not automatically count as failing to fulfill the deal.
For more stringent requirements, the city is counting on what it calls
Service Level Agreements, or SLAs. These agreements will probably end
up specifying things like how fast Earthlink will respond to reports of
broken equipment, or how many wi-fi nodes per mile they’ll need to
install. The city feels that by negotiating these agreements over the
life of the contract, they’ll be able to react to changes that they
can’t anticipate now. The downside of this approach is that the quality
of the network for most San Franciscans will be dependent on the
quality of these smaller agreements the city can negotiate. The Board
of Supervisors is being asked to approve the contract without knowing
the content or range of these SLAs.
There are better alternatives. One possibility is that the city could
build a network itself and lease connections to companies wanting to
provide the actual individual accounts to residents. The Board of
Supervisors Budget Analyst released a report concluding that it may be
fiscally feasible to build a municipally owned wireless network -- one
that would likely have significantly faster speeds than the proposed
one. A study was just released Friday that concludes the city could
build on its existing fiber optic installations to provide the backbone
of a truly 21st century network, one that could in the short term
provide faster wi-fi to every neighborhood in San Francisco and in the
long term provide up to 100 times the speed that Earthlink is
proposing.
The proposed agreement with Earthlink would be a possible solution to
the problem of giving already-connected tech-savvy people a way to
extend their access to the internet outside of work and the home.
That’s not the problem it purports to solve, however, and it will
likely fail to solve its stated goal -- which is to bridge the digital
divide by providing the platform to provide every San Franciscan the
opportunities that the internet makes possible.
Sasha Magee is an activist who writes at Left in SF
|
News Updates 