Quote from: pine on July 19, 2012, 09:23 pmQuote from: LouisCyphre on July 19, 2012, 07:15 pmInteresting idea, but why choose Java for it? It's an ugly, horrible piece of shit of a language and, contrary to Sun's/Oracle's claims, is pretty bloody far from platform independent.Agreed, at least for this purpose. I'm not so worried about being platform independent as I am security and speed. Java is simply too inefficient, I mean, this is a piece of software that's going to be doing a helluva lot of CPU intensive cryptography, if it were possible it ought be written in assembly.Exactly. The problem with using Assembly is the changes needed for every chipset. Far too much effort and there's already languages one level higher to cover that.Quote from: pine on July 19, 2012, 09:23 pmIn practice though, I'd like to see it written in something like lisp (piece by piece in an agile fashion, to get it off the ground quickly and iron out the logical bugs) and then translated and made more resource economical by a related working group into C over time which are back by a few versions. I was thinking something similar, except what popped into my head was rapid development in Python and then optimisation of relevant components in C. Mind you, I've got nothing against Lisp either (how could I considering I'm typing this in Emacs).Python has one other advantage with being able to happily work with numerous other languages. If the project is designed in a modular fashion it is good for tying that all together.Quote from: pine on July 19, 2012, 09:23 pmMight sound like doubling the effort, Not really, it's a solid development method.Quote from: pine on July 19, 2012, 09:23 pmbut there's a case to be made for an inefficient lisp working edition because you often want changes e.g. bug fixes to be rapidly incorporated into everybody's copy on the network ASAP and it makes sense from a 'two system' throwaway prototype development model too. Then while that's going on you have guys beavering away at making it more efficient in C until eventually it's ridiculously efficient and becomes the Gold Standard for lots of people to use, not just SR and related black market hidden services on the darknet, and that will give us yet more cover + feedback from a larger audience until the vast majority of serious bugs are found etc. Also, some important parts of the software should be written using rigorous methods like formal methods or something similar, so that we know the core of the system is literally proven to work.Yep, that all makes perfect sense.Quote from: pine on July 19, 2012, 09:23 pmProblem is getting enough of the right people with the right skillsets in one place at the same time for long enough to pull off a decent alpha version, and then continue for years with a smaller agile team. It is possible, I mean if Linus can use the Internet to create an entire operating system, this should be simple. Of course Linus didn't have to worry about the FBI putting a dent in the front door, but there are similar constraints.The trick here is to find a case for making such a project publicly. Identify all the requirements that a darknet black market site has and then identify a situation in which those same requirements would be needed that would not result in being hunted by the FBI, DEA or whoever.Essentially you're looking for a situation where people need robust anonymous and secure communications. So go through a list of human rights situations and pick one. A good example would be designing a system to allow the people of Mexico to communicate freely with each other without being identified and killed by the cartels. I know that some groups (including parts of Anonymous and others) have already been looking into doing something in relation to Mexico. Another good example would be something that could be deployed by dissidents in China. It should't take long to find plenty more use case scenarios.The real bonus is that if such a system can be developed then it really can help both the selfish goal of facilitating black markets and the altruistic goals of furthering human rights and protecting lives. I've got to say that while I like the former, the latter would be nothing short of pure awesome.