QUOTE(Infinium @ May 25 2007, 01:51 AM)

Those differences on the same Kernal can be seen as the unique identifier. The comparism with the other Kernal helps to identify the position of this information bit. Given that this identifier follows certain pattern it should be possible to adjust a generic kernal to it's own mobo.
You will find difference but there will be no pattern. Those sections of the NAND that are encrypted per-box are done so using the fuseset values. Thus, there be sectors that are encrypted differently per box and you can figure that out but you will not be able to 'adjust' a generic kernel to another motherboard without the fuseset values. The only way to get these values currently is to have an exploitable kernel. If you have already upgraded to 4552 or later, you can't get these values. So, there will be no downgrading at this point. All of the encryption of the NAND is pretty much figured out, you just need to read the proper threads.
For those hoping to downgrade, if you already have 4552, you can't go back to a kernel prior to it at this point. I am not sure about 5759. They may have done the same as the 4552 upgrade, which you can't corrupt the applied patches and downgrade.
As for the user who asked about cutting traces, etc... The NAND has a pinout on the bottom of the motherboard. Very easy to solder too if you can solder.
QUOTE(The Prankster @ May 25 2007, 02:20 AM)

That poses quite an interesting view. Would be cool if it was plausible, though it's all signed by M$, thus not allowing any altering, so if we removed the per-box key, and replaced with another, it would not be signed and would not work and not bootup.
If you could reverse the encryption of the per-box sections, had your fuseset values, you could in theory re-encrypt those sections and have it boot. No need for it to be signed.
Caster.