6 ms·
1) Just a funny fact, the Craig Wright who's first author on this paper is that Craig Wright. 2) This is a good paper on the situation, but as I've mentioned b
by jcrawfordor 5y ago
1) Just a funny fact, the Craig Wright who's first author on this paper is that Craig Wright.
2) This is a good paper on the situation, but as I've mentioned before tends to miss the point. The data remanence concern today is usually non-volatile caches and remapped sections of the media which are not documented or accessible to the host. This problem is best known for SSDs but very much exists on modern platter drives as well. No number of host-based overwrites will reliably overwrite these even once, and the manufacturers do not document or disclose the behavior.
- anfractuosity 5y agoDo you mean a type of spinning disk that also has flash memory for caching? I thought spinning disks normally used volatile memory for the cache?
- jcrawfordor 5y agoNo, platter drives usually don't have nonvolatile cache although it's hard to say for sure. Enterprise drives are more likely to. The bigger issue with spinning drives is that they move data around the platters in ways not disclosed by the manufacturers, including both optimizations (writing near heads) and error corrections like ending use of problematic parts of the platter. These all leave bits of data in "non-addressable" parts of the platter.
- anfractuosity 5y agoAh, that's interesting. I wonder how hard it'd be to find these areas of the disk, I wonder if the location of them is stored in some kind of non volatile memory, so the disk can easily find them.
- dzdt 5y agoAnd the Dave Kleiman who is second author on this paper is that Dave Kleiman, the one who most likely invented bitcoin. Wright has claimed to be Satoshi Nakamoto but has proven himself to be a liar on this topic; by his actions he clearly believes that Kleiman controlled the Satoshi bitcoin fortune before his untimely death.
- nullc 5y agoDave Kleiman had absolutely nothing to do with the creation of Bitcoin. Prior to his death Kleiman was IT support staff for a small county sheriff's department in Florida. He has no relevant expertise -- e.g. the only evidence of Kleiman ever writing any program ever was a simple visual basic script to automate checking the windows registry for forensic information. Literally the only thing remotely suggesting any connection is documents from Wright which are provably forged and created after Dave's death. [Wright's forgeries tend to be extremely bad: Stuff like changing the year in emails but not fixing the day of the week and misspelling Dave's name -- in emails supposedly from Dave. Accidentally producing unmodified originals, citing laws that didn't exist until years later, using software versions that didn't exist until years later, referencing URLs on his own domain names that he didn't register until years later, accidentally producing "dave"'s private keys in discovery ... but they're also extremely numerous, so it's easy to get smothered under the total mass of them.] Wright invoked his former friend in his scheme posthumously to solve a specific logistical problem: Wright himself can't program (certainly not in C/C++). So in his forgeries he included someone who couldn't contradict his stories so that Wright could beg off his inability to answer questions as "Dave did that part". Fortunately for Wright, it turned out that none of the eligible victims asked any hard questions. Unfortunately for Wright, Dave's family came after him for their share of the fortune Wright claimed they created together. Fortunately for Wright, Dave didn't actually help him (because neither had early involvement in Bitcoin). Unfortunately for Wright, all the easily proved forgeries in the case with the estate end up making it look like Wright is covering up the involvement to hide what he owes. Wright's frogurt is cursed, but also doesn't exist. (So that's good?) Kleiman isn't the only deceased person that Wright involuntarily enlisted into his con-- Wright as also claimed that he was aided by Gareth Williams (some MI6 spy that was in the news at the time Wright started spinning this yarn) and David Rees -- an older academic who was disabled and unresponsive in a care home at the time Wright claimed to have been working with him.
- oudekaas 5y agoyou sure take your time to discredit Craig don’t you Greggles…… For those reading Greg Maxwell’s rant (founder of Blockstream/ hijacker of Bitcoin), just be careful using someone with obsessive compulsive harrassive behaviour as a trustworthy source, especially when there is financial bias involved. Greg Maxwell has been harrassing Craig for years now, he is obsessed by Craig and BSV, and wants to see it fail for obvious reasons, if Craig turns out to be Satoshi (hint: he is……) and proves this in court (yes Craig is dying to prove it in court) it will highlight the massive consumer fraud, Blockstream is complicit in. You may want to figure out why Greg is so obsessed about BSV, maybe just maybe…. BSV is the original Bitcoin design that allows nanotx and has been shown to be capable of handling large scale transactions, defeating any purpose for BTC and ETH, and literally 99% of the crypto space to exist. Bitcoin does all, BTC does Jack! This is not investment advise, this is about highlighting fraud (Bitcoin’s rules were set in stone…… by Satoshi, like monopoly most people tend to play by the rules. Not Greggles though, he prefers to see a Bitcoin being played by his rules……..
- sharikous 5y agoWould that mean that, hypothetically, a bad actor could convince the manufacturer to place sensitive data about you in inaccessible parts of the disk? So short of destroying it completely, a computer (especially one where permanent storage is soldered in) could be assumed to contain anything that was stored on it even once? Well when the government will become very interested in hoarding old electronics and registering every facility able to destroy them, we should become suspicious. I would be very relieved to be proven paranoid.
- nullc 5y agoYou don't even need a bad actor, the drive will do that on its own. Malicious firmware only needs to be invoked if the attacker wants to be SURE that the preserved data 'just happened' to cover e.g. your crypto keys.