Brewster Kahle’s Internet Archive is donating bandwidth (via fiber) to San Francisco housing projects, the first of which to be connected is Valencia Gardens. Residents will get fast, free Internet access. Actually the project is spearheaded by Ralf Muehlen (SFLAN, which is an initiative of the Internet Archive), Tim Pozar (United Layer), Michael McCarthy, Bruce Wolfe, etc. All have been critical about San Francisco’s deal with EarthLink, which is now dead.
For those of you with short memories, recall that San Francisco issued an RFI in 2005 for a citywide wireless broadband network. The purpose of the RFI was to gather responses so that the city could draft an RFP.
I posted the responses – two of which I found compelling: from the Bay Area Wireless Research Network (Tim Pozar) and SFLAN (Ralf Muehlen) – in October 11 and 14, 2005, respectively. Those are the people who are now bringing high-speed broadband to SF’s housing projects, and hopefully to more of San Francisco.
I am copying the articles I posted in 2005 in their entirety below, to make it easier for you to read.
(1) Bay Area Wireless Research Network’s plan for SF
BARWN proposes:
1. A municipal broadband network will need to use a mix of different technologies such as 802.11 (WiFI), 802.16 (WiMax), copper and fiber to address the different needs of deployment such as geography, rights-of-way, etc.
2. The City should require that the network comply with open standards established by recognized standards committees (ie. IEEE, IETF, etc.). The network should not be dependent on proprietary protocols or APIs.
3. The City should provide a Level 1 / 2 high-speed network (i.e. Fiber) as a backbone for paid access to ISPs and clients in San Francisco.
4. The City should work directly with business and community groups to identify broadband needs.
What I found most compelling about their response
I like BARWN’s proposal that the City build and maintain the physical layer (Layer 1) which is a fiber infrastructure and that the data link layer (Layer 2) be managed by a non-profit consortium of network stakeholders. On the network layer (Layer 3), private companies can offer Internet access, VPNs, or “anything that can support either dark fiber or lit fiber using Ethernet over fiber protocols like 802.3ae.”
BARWN’s model ensures that no commercial service provider – SBC or Comcast – owns the network and has the power to deny access to competitors. In essence, BARWN is asking that the city decouple the owner of the infrastructure layer from the owners of the businesses that actually provide access to end users. In Europe, this decoupling, achieved by the regulators forcing incumbents to wholesale access cheaply to competitors, has led to massive competition among Internet service providers (20 EUR/mo for 20 Mbps in Sweden, 10 EUR/mo for 10 Mbps in France, and so on).
Why I like fiber: bandwidth. As Tim and David point out in their response, the incumbents have not provided the bandwidth that users require:
Applications that use the Internet are becoming more bandwidth hungry. Multimedia files are becoming larger, the media incumbents are understanding that their audience is getting their content on the Internet. Voice over Internet Protocol (VoIP) requires symmetrical broadband access, unlike the broadband offerings that the cable and DSL providers offer now. SBC’s residential offering is limited to 1.5Mb/s down and (up to recently) 128Kb/s up. Comcast’s offering is also asymmetrical. This restricts many peer-to-peer applications such as VoIP, gaming and user publishing of content. Peer-to-peer applications are just starting to be developed. Demand for symmetrical high-
speed bandwidth will be ramping up at an exponential rate.
The commercial providers only want to stuff THEIR content down the network to your computer. They do not want users to create their own content, which will probably be better than the commercial providers’ garbage, and share it with other users. Not good for community building.
How do you share video on slow US networks?
We are seeing new applications that allow users to upload video and share it with other users. These are similar to Flickr’s photo sharing site but for video clips. YouTube and Vimeo are only two examples. I know of several people who are working on beta versions of video apps.
My question has always been this: how are Americans going to use these applications when most US service providers offer low bandwidth for upstream traffic? In Europe and Asia, a lot of providers are offering 10 Mbps and higher bandwidth for uploading. My advice to US video applications developers is to set up foreign language versions of their websites because a large number of users will come from other countries.
Read BARWN’s response
I encourage everyone to read BARWN’s response. It’s short and very well written. To read BARWN’s response, download the PDF from here.
(2) SFLAN’s plan for San Francisco
SFLan is a Community Wireless Network in San Francisco. It is a project of the Internet Archive founded as a non-profit in 1996. The Internet Archive is headed by Brewster Kahle, Digital Librarian; SFLan is headed by Ralf Muehlen. According to SFLan’s documents, it is the largest outdoor Wi-Fi network in the city and has been providing free access since 1998.
SFLan proposes a wireless broadband network acting as a city-wide Internet Exchange, with ISPs and CSPs providing value-added services. The network should cover mostly roof-tops with indoor and outdoor CPEs enabling indoor access. To make the network survive disasters, and be useful in such a situation, it should utilize mesh technologies on the backbone, and integrate with photovoltaic and in city generation systems.
What do I like about this proposal?
1. Network neutrality
2. A solution based on open standards, Open Source Software, and industry-standard
hardware to avoid vendor lock-in
3. No captive portals
4. Anonymous speech
5. Basic access is free of charge
6. Emergency plan: to ensure the network survives a disaster even at reduced capacity, a “subset of nodes will use meshing technology on their backbone link, some of which might utilize the 4.9 GHz band. These nodes should have solar-powered battery-buffered power supply, integrated with the City’s ongoing photovoltaic and in-city generation plans. Even if all Internet uplinks are cut from the network, it could still serve as a local network, assisting in situational awareness for first responders.”
7. Cost of building the core network financed by a bond measure or the general fund; CPEs to be purchased by private households (with exceptions for low-income families): the city funding and owning the core network would avoid having one company own the core network and discriminate against other service and content providers (as in the cable franchise model).
To read the entire SFLan proposal, click here to download the PDF.
Of course the NYT is not patting all of the correct persons on the back here and missing the point a bit. Brewster is donating the bandwidth. He gets bandwidth for less than $10 a Mb/s. If we assume that Valencia Gardens uses all 100Mb/s then he is donating about $800 a month of bandwidth. Brewster certainly should be recognized for that. But there are a number of other players that play as big of roles in this turn up.
Much of this work in having this happen was the result of Michael McCarthy who was the point person in getting the Alice Griffith housing project lit. Michael is now with DTIS/City of SF. He is the most clued in the City of how to get bandwidth into areas that can’t get broadband and try to address this digital divide.
On top of that there are other players in this effort such as UnitedLayer (the company I am part-owner of) who is supporting some of the fiber infrastructure and cross connects (and who also donates the bandwidth to the Alice Griffith Housing Development), Internet Systems Consortium who is providing some back haul to cross connect Internet Archive (IA) to the City’s fiber network. Last but not least is Ralf Muehlen who has been leading up the SFLan effort for a number of years now and was the one that brought IA into this project.
I should also add that getting to this point where the City is open to this kind of deployment has been a long road. Ralf, Matt Peterson, Bruce Wolfe, I and others have been showing up at various committees and supervisors to get them to understand the idea of municipal broadband. Brewster’s effort has always been to have the private individual to deploy for others (SFLAN). Leveraging muni-owned infrastructure has been a concept that I and others have been pushing for the last 4 years with the City.
What is really interesting about this deployment is not that fact that Brewster is donating bandwidth, but the fact that the city can deploy and own the fiber and have anyone use this pipe for any purpose. The last mile is the hard part in getting folks connected. The city should be owning the pipe much like they build the roads for folks to use. No other company can do what the city can.