These are notes on various aspects of making XC a real company. This is intended to provide a very general outline, without getting bogged down in a lot of detail.
The initial step is to agree on a CEO, to direct the formation of XC. CEO candidates should present a plan to recruit a management team, and first-year financial plans, as well as any proposed emendations to this business plan. It's not clear how to decide on a CEO, but a public forum of proposals would be a reasonable step. (Of course, you could just start your own company, fait accompli, were it not for the need to garner support from the community.)
There are two basic approaches to bootstrapping, depending on how much capital one might raise (and how soon):
Some mix of these approaches can be synthesized, depending on how much capital is available, how much business is being raised, and what one chooses to emphasize. Moreover, the long list of services to be offered can be phased in (e.g., 24x7 backup is not a day-one priority).
One piece that is something of a priority is to come up with a configuration for the gateway computer. One proposal is to use a rackmount computer, small rack, and UPS. This has the advantage of looking like something more substantial than a PC, but most currently available equipment along these lines is prohibitively expensive. We may wish to find a hardware vendor and work out a semi-custom design that better meets our requirements.
We may also want to form limited time vendor-exclusive relationships for things like structured wiring boxes and fixtures, and home automation devices. The guiding principle here is to get to the best available technology as fast as possible. In the long term, vendor importance is a key principle, but in the short term it could be more effective to limit our options in order to simplify our systems and get more and better vendor input.
There is already a huge amount of open source software that covers many of the most important requirements for home networking. Of particular import is the Linux operating system, including the networking support needed for firewalling and routing; Apache for serving web pages; Samba for Microsoft-compatible file- and printer-sharing; database managers; backup programs. There is VPN support, for secure connections between home and office, and potentially for access and monitoring while the user is away from home. There is package management software for updating the system.
While the gateway/server computer runs Linux, the end-user's own computers can be whatever the user fancies. The principal user interface to the system can be provided through web pages, CGI, PHP, and possibly Java, which can be accessed from any authorized machine with a web browser. The design of the user interface needs to be worked out, but the tools for implementing it already exist.
There are also open source projects for home automation, such as LinuxHA and Misterhouse. At present, these projects appear to be most interesting to DIY users. We need to assess them, and work towards making them more intuitive and robust for ordinary users.
Open source projects to date mostly concentrate on device support, networking protocols, and tools. We've scarcely begun to imagine the software that will flesh out the home computer network into something that becomes an indispensible luxury of everyday life. What is clear, however, is that this software must be trustworthy, reliable, robust, easy to understand and use, predictable, and fundamentally useful.