latest comments:
currently, several companies 'own' the Internet, through the dedication of lines and routers that enable the global community to go from one ip address to another. unfortunately, these same companies are not always looking out for the interests of the average web-surfer. through in an unhealthy mix of cable, dsl, dial-up and satellite providers, and you've got a recipe for disaster: too many people trying to make money selling slow service to people who don't know better. the solution? an open public network, utilizing redundancy and speed to enable a large volunteer community to run their own network. by putting the network back in the hands of the people, the internet truly becomes free for all.
the technique? simple. instead of relying upon isps to allow access to the internet, individual users purchase small wireless access points to tie into a larger grid of more dedicated volunteers with small routers/switches who construct a wireless/wired grid of interconnected nodes: the vbone, a volunteer backbone. the vbone itself will be open to all, and will support any traffic without limitation. individual members of the vbone may use QoS or packet filtering to reduce load through their neighborhood, but the vbone itself should be able to route users to their desired locations in the most opportune manner. if a single vbone node goes down, it should be easy for the network fabric to adjust itself and for convergence to quickly begin to route traffic around the problem. with enough nodes, large congestion points will be avoided and the overall speed and usability of the net will increase. plus, the less people who strain the existing system, the easier everyone's life will be.
by incorporating the latest technology into the wireless access points, as well as the fabric of the vbone itself, we hope to prevent the network stagnation that currently plagues dsl, cable and satellite customers. not satisfied with your current connection? simply purchase a new access point, or download some optimized software, to immediately see a speed difference. unlike computers, users are used to seeing only incremental increases in network speeds, from 2400 to 4800, even up to DSL speed. once the t1 line becomes a slow pipe, then the vbone will have succeeded.
edward is just your average, run-of-the-mill, white-hat nobody. got it? that's not what he'll always be, however.
edward occasionally refers to himself in the third person, bob dole style, but frequently uses his real name. his life has been a fairly smooth upward trajectory, with frequent bursts of amazing luck tempered by periods of just better-than-average luck. he has never been seriously in debt. he has overcome a few major illnesses & some operations along the line. edward appreciates the power of morphine.
edward built this site, and refines it from time to time. the blog is the only thing that stays consistently up-to-date. edward doesn't review as much as he'd like to, but he does love giving advice. he was annoyed once when he had to take down a particularly interesting column.
edward is not a style nazi, but his bible of choice is probably the AP stylebook, which is the sliderule of the 21st century: it can do everything. edward actually doesn't know what a sliderule can do, and has only seen one once. that's what calculators are for, right?
edward loves politics. that's why there's so much political stuff in the blog entries. go figure.
as a frequent user of microsoft products, it saddens me to see some of the free software movement bludgeoned by those in the corporate/government world for lack of solid enterprise features. considering that even Microsoft is a relative newbie when it comes to a true directory service, this is disheartening. after working with active directory for about two years, i think i've come to appreciate its strengths and deride its weaknesses, particularly in respect to other microsoft products that use/ignore it while duplicating its effort, like SMS. this project will be dedicated towards developing a free software solution that not only can perform as well as active directory or nds, but far better, without relying upon a single OS the way active directory does. nds is superior, yet even it is not open. gnads needs to meet several goals in order to surpass active directory and nds. it must be easy to install on a variety of platforms; it must scale to over ten thousand resources with no problems; it must be distributed, and there must be no single point of failure; it must be extensible, so that future alterations to the directory schema may be easily made; it must incorporate as much information as possible about each resource, yet not overburden queries with such information; it must be fast, for both searches and regular access; it must provide a secure means of authentication. these goals are not easily met: ms and novell spent years assembling their respective directory services. hopefully, the inclusion of many developers will help speed this project along so that it can displace Active directory before it becomes a de facto standard. ideally, a gnads client/server system could be setup in a few minutes; the client would poll the particular workstation it was installed on, adapting itself to the particular environment it was in, and reporting the environment back to the directory itself: thus, features which are often only done through add-ons to NDS or active directory, like inventory management or user access tracking, could be incorporated into the directory itself! if i had a small office containing two macs, three win9x machines, two win2k machines and a linux and netbsd box, (plus a linux server for gnads!), i could install the client on each machine, point it to the local gnads server, and instantly watch the directory become populated with information about each workstation. after adding a list of users and groups into the gnads server, users could log onto each box and access features of the directory; as each user logged in, that information would also be added to the directory. through gnads then (in this hypothetical scenario) one could access a list of programs installed on each system, monitor a user's usage, etc. it would quickly become the end point around which all the services were oriented. this is the goal of gnads: ambitious, but possible.

