again what if your dumps spit out the same garbage at the same point? is it good data? dont think so. your logic is not sound.
i reiterate my point that if you dont get 100% matching dumps you should not proceed
Please explain, logically (use examples if you need to), where the flaw is in my logic.
Meanwhile, allow me to explain the error in your argument.
You are assuming that if you do two reads and they match they must be good reads. This is not really true, you may have some odd scenario that causes a bad read, where that bad read happens in the exactly the same way every time. I.e. you read block 0 twice, you are given the same data twice, but that data is wrong (assume then, for completeness, that the rest of the chips reads fine). So you now have two matching dumps, which based on what you are saying means you have a good dump, right? Obviously not, so clearly your premise is incorrect, however you and everyone else (me included) accepts this premise. And luckily I can't really imagine a scenario that would do that, and that's why people who just do two dump that match generally don't come back complaining.
Now, this is the exact same premise that this tool works on. So although in theory it is flawed, in practise it's good and you, me and everyone else is happy with it.
When you have your two matching nand reads (done in full, your way), if you do a third and it doesn't match (say it has one random byte different in the whole 16mb image) what do you say? That third image must have had a read error, right? That can't be the right one and the other two wrong, because the other two matched. And what are the chances of two wrong ones matching? That's the logic you are using in saying that two matching nands must be right.
The only additional assumption that this tool makes is that you don't need to work with the entire nand image at one time. If you read the first 8mb of the nand twice and they match, then you read the last 8mb of the nand twice and they match, why can you not join the first half and last half together and make a full nand? No good reason why you can't. Suppose you make a third read of the one half of the nand, but this time it has a byte different? You'll conclude this one was the read error. The other two were the correct ones, because they matched.
This is all the tool does, but it does it at the block level which is smaller. The risk of a purely coincidental matched misread does increase with the smaller unit of comparison, but in a 16k byte block the chances of this are a lot worse than winning the lottery. If it's a non-coincidental match, from that theoretical scenario we can't think up, it'll hit you as bad as it hits me.
trying to reconstruct a good nand from a pile of dung is pointless
Luckily that's not what the tool is for. Well, I have no control over what data you give it, you can give it 3 pieces of dung if you like, but it won't be able to turn that into a complete image. Unless you give it 3 identical pieces of dung, in which case it'll quite happily reform them into a new piece of dung. Of course if they're identical they'll already have passed your test of their quality, so using this tool leaves you in no worse a position. Then of course there is the instruction to sanity check the produced image in something like 360flashtool.
and are you then going to be held responsible for that? love to see you firing out replacement xboxes coz ur tool made some1 think they could proceed with a bad setup.
I'm no more likely to do that than you are after telling people that matching dumps are what you are looking for. If this tool will fail it will be for the exact same people your test fails for.
theres no need to be so defensive you can get critisism as well as praise here you know.
I'm explaining. If that's being defensive wouldn't you have to concede that an unprovoked attack against something using an invalid argument is aggressive (weapons of mass destruction, anyone?).
i would also urge any1 reading this to look at their setup before they assume your app will be their saviour.
Yes, that's a good idea. A properly working setup will save you time and effort. However there are many reasons why a setup may not be perfect, not all of which are fixable (I have one LTP port in my house, it appears to suffer some minor timing issue or such like, preventing it getting consistent reads of a whole 16mb in one go), that's why this tool was written. More serious errors will not get you as far as this tool or this tool will not give you anything claiming to be a good image.
Dilber MC Theme by HarzeM