January 2005 note - I haven't done much with this in a long time. This data is fairly out of date. This project languished as there was lots of initial excitement but no follow-through in the Boston area. Even the media frenzy about free wireless networks over 802.11b has ebbed. Oh well. But the web logs still show a lot of hits on these pages, so I guess they are serving some purpose, and so they will remain up, but won't be frequently updated.
The philosophy of this project is that the only way important and often controversial information and opinions can continue to be spread is if the communications infrastructure over which such data is transmitted is as free from interference from government and commercial control as possible. The Internet has given us wonderful power to widely disseminate thought provoking material at little cost. Unfortunately, there has been some significant efforts by governments and others to curtail this freedom of speech with such tools as the Communications Decency Act. Though that act and other similar methods have been successfully challenged, activism in public policy is not alone sufficient to guard against the encroachment and eventual eradication of that which has been achieved. The physical infrastructure must be controlled by individuals and small, dedicated groups.
Towards this end, we conceive of a network free of commercial Internet service providers, governments, telephone companies, and ad-hoc Internet regulatory bodies of dubious legitimacy (such as ARIN and Internic). By use of license-free (so-called "part 15") and Amateur Radio systems to transmit data, we eliminate interference by telecommunications companies and the various bureaucracies that regulate and promote them. While there is some cost and technological sophistication and attention required of the operators, all efforts are made to keep the costs low to encourage participation of a spectrum of operators without the necessity to charge money for service in order to recoup costs.
Network design has been kept patch-work on purpose. We feel that the use of surplus, home-built, and commercial gear where available and appropriate will keep costs reasonable. The current plan is a two-pronged approach, the first utilizing wireless ethernet cards (802.11 and other types) for high speed, short distance, line-of-sight links, and VHF/UHF Amateur Packet for lower speed, longer distance hops. This network is in implementation stages in Boston, and in planning stages in New York and San Francisco.
Sinister Networks is offering a portion of its Internet bandwidth as a gateway. We foresee several links to other networks: Internet, traditional Amateur Radio, telephone, and other gateways being put in place where and when they do not interfere with the network philosophy or operation and maintained only so long as they don't.
Low speed transcontinental and intercontinental links running on HF radio bands carrying only priority traffic should appear, but hopefully become less restrictive as additional capacity is added. Until this implementation, data will be carried between the disconnected portions of the wireless network via Internet, or commercial landline facilities.
Security is an obvious concern over a wireless network, most importantly encryption. There will be several layers of encryption on the network. Many links will be utilizing Direct Sequence Spread Spectrum. While this does provide some privacy, we do not feel it is secure enough, particularly in the WaveLan implementation, which only has 2^16 possible keys. To provide better security we feel that some link-level encryption is called for. One popular candidate for this layer is IPv6, which includes encryption natively, and is well supported in various operating systems. However, the very fact that a standards body has approved it suggests that the NSA, which has a tradition of sticking its fingers in all encryption issues, has weakened the protocol. To keep the spooks at bay, still another layer of encryption may be used on the session-level, such as the triple-DES protocol of the ssh program. However, this would be a completely voluntary standard, the user is free to forego it. We intend to keep up to date on encryption and other security issues to maintain the viability of the network.
Linux and other "free" operating systems will be the main development platforms for the project. The use and promotion of these OSes fits in well with our philosophy. Besides, they are reliable and easy to write programs for.