A year or two ago I had this idea that's related to our discussions... In short, it was to create a open source platform for WISP use. I called it WISP-OS. All the functions of routing, firewalling, dhcp client and server, and all the other networking functions are out there and consistently being improved in the open source community. What, however, is needed is not another implementation of routing or firewalls, but deep down fundamental efforts to improve the drivers for the common, cheap chipsets.
I got several interested parties including developers and WISP's, but the obstacle is the funding. The reason you need substantial funding: The wireless driver holds the key here. You need the license from Atheros, and that alone is a serious chunk of money. We came up with a couple of viable methods of making the idea work. The driver development has to keep the Atheros sources closed, and like other people have done, fundamental adjustment of the MAC would be the ultimate function. I saw this coming down the road when then software companies were moving toward becoming a closed hardware/software platform. The idea was to produce a licensable driver that could be integrated into any new hardware that might come down the pike, and put research into development of features that could be universally shared. Right now, each developer has created their own 'non interoperable' feature set. So you want WiMax? Great. Only the basic feature set is interoperable among all. Anyway, the purpose was to let WISP's guide the direction of development. So, you want to use the cheap NanoStation? No problem. The open source community has almost everything needed. And each hardware platform could have any/all advanced features. So, instead of Star-OS having great performance, but only with itself, same with MikroTik, any hardware platform could share a full feature set. This would require considerable funding, to do this development, but that funding would be obtained from per-unit licensing scheme... something not expensive per unit. Also, since it mostly would be comprised of open source software, the development for each new board or cpu could be done by individuals or even small g roups or companies, and only the licensed, closed wireless driver would have to be "paid" development. The initial cost for this could be born by 50 wisp's and be relatively small. The largest initial obstacle is the Atheros license cost... But, this would spur movement toward much greater interoperability - or at least the possibility of greater interoperability. So, while each hardware platform developer is re-inventing the wheel... It would no longer need to be done...simply license a great set of features that are driven by the WISP's who guide the development... As WiMax modules become more available, the same kind of driver/licensing system could be done for it, too. The same economies of scale and competitive production could apply to WIMAX as they have done for the 802.11 platform - specifically Atheros... This empowers individual wisp's to become legal integrators, like the modular fcc approval has done for Star-OS and others. Like I said, this idea is an old one for me, one I gave up on because nobody seemed to be interested, but it IS a viable notion and if this had been started back then, it would now be the key solution to much of the consternation now . ++++++++++++++++++++++++++++++++ <insert witty tagline here> -------------------------------------------------------------------------------- WISPA Wants You! Join today! http://signup.wispa.org/ -------------------------------------------------------------------------------- WISPA Wireless List: wireless@wispa.org Subscribe/Unsubscribe: http://lists.wispa.org/mailman/listinfo/wireless Archives: http://lists.wispa.org/pipermail/wireless/
