Bregalad wrote:
tokumaru wrote:
Well, there are only so many cartridges produced by Nintendo, and they will eventually run out. We can't rely on them forever. I guess for personal dev'ing it's ok, we can always bust a cart or two, but for "mass" production... I don't know.
I agree. I'm strongly against wasting cartridge to destroy them if it is not for great use of them. This should be avoided as much as possible.
Producting two slot adaptater wouldn't be hard, would it be ? It so, the CIC have to be reverse enginered to found the most accurate way to disable it. Or simply explain to people how cut their pin 4 of their CIC.
The way I see it, you still need a cart case to make carts, so you need to use production carts anyways... so remove the lockout chip while you're at it, and re-use the case too.
Making some kind of funky two-way adaptor will be super duper expensive, as will new cart cases. You're looking at $10K or more for custom molds. This is way outside the realm of possibility for most hobbiests.
Also, most carts will be a run of 100 carts or less. This is a drop in the bucket. There's thousands and thousands of SMB/DH and SMD/DH/WTM carts kicking around that no one wants. And contrary to popular belief, it is easy to re-use the glop top lockout chip, just use some wire cutters and cut the board off leaving just the lockout chip. You can then solder it to the edge of your board with 6 connections. No muss, no fuss.
Anyways, better people than us have tried to defeat the lockout circuit, and every one of them has failed. Atari was on the cusp of victory, however, when their lawyers snagged the 10NES code... making alot of the work they did moot.
Today we have some fairly powerful computing at our fingertips, so some kind of brute force attack *might* be possible. I can record gigs and gigs of data if someone wants it, but without someone to process it and apply some good cryptographic knowledge to it, we're sunk.
I highly doubt if I dump 20 gigs of data (using my simple delta modulation scheme to pack the data down 300 fold or so) I seriously doubt we will see a repeat of the data stream. I could set an FPGA up to look for a data stream repeat, but IMO I think we'll be waiting months or more before it repeats. If it's more than 64 or so bits, we won't see it repeating in our lifetimes probably.
Again, I could set an FPGA up to dump the data in realtime, or possibly faster than realtime, but I need somewhere to stuff it to make it workable. If I fill up a 400 gig HD with nothing but lockout chip data, it may not tell us anything useful even then.
I am fairly serious about cracking the chip, but I'm only good for the "back end" of the affair, that is, supplying the data for others to work on. I have absolutely no experience cracking any cryptographic functions.
I'm sure if we show the right people some frequency distribution patterns of the data (which have been generated before) they MIGHT be able to shed at least a little light on the situation. A true "cryptographically secure" algo would not have the curious frequency distrubution we have seen, and it would instead have a very even one. Our data definitely does not look like a gaussian noise source, so there are some very very important clues that I just don't know how to interpret.
Anyone got clues who we can talk to to get the ball rolling on some good crypto doods?