The Others wrote:Apart from the crappy live cd idea, I think the best version of NS ever got designed. Maybe April Fool's is the perfect think tank solution.
LOL.. When writing the readme file, and trying to come up with believable B.S. G8t and I both started to drool.. Because "What If" it was possible ?
As I understands it, one of the primary reasons that source code isn't available, is because the program contains a lot of code that was developed under Non-Disclosure Agreements. I would guess that it is the code that deals with the direct firmware interaction of the Hermes chipset and the chipsets partly supported, such as Atheros and Prism.
Theorethical a Linux port of NS, wouldn't need those specific parts of the code, if they could utilize the existing Linux drivers, which also would enable a passive scanning ability, like Kismet has.
In effect modularizing the code, would get by that hurdle, and possibly also enabling a Plug-in system such as the one described in the NS.0.4.1.AF.2005 readme.
The question is if we could get Marius interested in such an idea, as there would have to be two different codebases, one for *nix and one for the Windows platform. I for one, would be pleased to contribute on the *nix side (alas the possibility of me actually doing so, are non-existant).
If you look at some of the ideas behind the New Core Kismet development, you can see that Mike actually has thought along those lines, by wanting to modularize the program.
I wouldn't be surprised if, at the end of the day, the New Core Kismet, also will incorporate the active scanning which NS utilizes, and maybe some of the featureset described in the NS 0.4.1.AF.2005 readme.
Convergence is not just the name of the game in hardware.
But if people really want to see a netstumbler similar to the one in the AF event, nothing stands in the way of starting up a project making it come true. Afterall, I don't think MacStumbler has any shared code from Netstumbler, so why should NixStumbler need to be based directly on Netstumblers code ?
We (as in "We, the people.." not "We, an already defined group of programmers..") know what we would like the program to do.
We know how we would like the GUI interface to look like. We have some ideas which might or might not be implementable..
There are knowledgeable people around who probably wouldn't mind advising and providing code snippets. There are Opensource programs such as Kismet and WaveMon, where one can gain information from the source, on how to do stuff.
It is just a question of getting it started and coded. But this is probably the base for a a new thread in the Linux subsection. Feel free to start it up.