Raven Darkheart said:
come on now
when the ds came out it took em at least a year to come up with a slot 1 solution that was stable
so we shouldnt panic yet
A lot of that time was spent just reverse engineering the handheld to understand how it worked. The Slot-2 solution was actually pretty crude, it didn't take very long because it wasn't based on any kind of advanced understanding of the hardware. Boot the DS with a real game, then throw it an instruction to a memory address that should not have worked if the DS firmware was written correctly in the first place (which is why Nintendo fixed it with later firmwares). The advantage now is that we have a good understand of the hardware and while the DSi changes some things, the specs of Slot-1 and the main CPU will not have changed. We know how to patch games on the fly to deal with the additional delays of NAND flash, and to intercept calls to read/write savegames. And if my understanding of modern flashcards is right, we even know Nintendo's original DS game signing key. It shouldn't take a year, the hardware has been developed, it'll be a matter of what mechanism is blocking current carts, and finding a vulnerability to bypass it. Nintendo has proven to be poor at correctly implementing security measures in the past.
It will be interesting to see if the speculation about a white-list are true. If Nintendo is smart they've added an additional game signing key (one we don't know) to new titles, and this is where the white-list would make sense. I can't imagine this holding up for very long though, it relies on the cart not lying. Since these are active devices, lying is very easy to do. We can't fake being a new game that the DS doesn't know about (due to new keys) but we could fake being an old one, albeit at the risk of starting an arms race with Nintendo in future firmware updates.
I would not be surprised if this new blocking mechanism made it in to the DS Lite eventually though. Clearly it's to Nintendo's advantage to break old flash carts in both the DSi and the older DS Lite.
On that note however, I sincerely doubt we'll see a SD hack on the DSi's SD card slot. What makes the DS unique is that it's meant to execute games coming from fast ROM memory, and while this isn't quite as literal as it was with the GBA (which had virtually no RAM because of the ROM), it still holds true. For SD flash carts, timing hacks had to be developed to trick the DS in to putting up with the fact that reads from NAND flash are not as fast as they are with the original ROM (If Openchip is still here, he could probably say a great deal more about this). This is part of the reason why every flash cart has a small processor on it, because it needs to constantly intercept things. Making the DSi read from its own SD slot means that we won't have that processor and we can't execute games simply by throwing the bootup routine a new address that points to the SD slot. The firmware itself would need to be modified to perform what the flash cart processor currently does, and I'm not sure that would work for any number of reasons. Not the least of which includes if we can even write custom firmwares, or if it's possible to shoehorn one of the ARM processors in to doing this without throwing off other timings or slowing down the DS. Getting current flash carts working on the DS will probably be far easier than going to the next level and executing games from the SD slot.
