Can you (or Tok) point to a part of a cryptocurrency which isn't cryptography?
the dev team, the website, any part of the wallet not doing a cryptographic function, the buyers, the masternode operators, the network transport, whatever isn't going into a cryptographic function.
This thread is moving pretty quickly (ten pages of mudslinging in half as many hours!) so I'm not sure you'll see this, but there's an important misunderstanding here. You're right that the human beings aren't cryptographic (e.g. the development team, market participants, etc) --- though it's important to observe that this makes them very unsuited to cryptographic functionality. gmaxwell had an elegant description of cryptography as "technology by which we hope to confine and constrain the nature of information" despite information respecting no ownership, borders or morality. He described this as "inherently subtle and fragile", which it is, but this indifference to political and social pressure also make it efficient (human trust is
expensive to build and maintain!) and robust against a lot of political and social pressures that human systems are not. This robustness is the cypherpunk motivation for bringing cryptography into everyday life: anything we have the technology to do cryptographically rather than socially ought to be, since social systems can change quickly and unjustly. Nowhere is this more true than finance, so cryptocurrency is a perfect environment for this kind of thing. So cryptocurrencies try to eliminate human decision points wherever possible, and where not they try to set things up so others can't override each others' decisions (hence the value implied by buzzwords like "decentralization" and "censorship resistance" and "public verifiability").
I'm not entirely clear on what masternodes do these days, but I infer that their actions affect users' privacy, i.e. they are "confining and constraining" information. This is a cryptographic function too. Human decisions may factor into their behaviour, but they are still performing a cryptographic function; human involvement does not change this, only changes the failure modes.
All this to say that basically the entirety of cryptocurrency really
is cryptography. Even many of the human parts. The network transport is part of the cryptosystem: it needs to be designed to prevent modification of data in transport, authentication of data even when endpoints are anonymous and spoofable, etc. A wallet is part of the cryptosystem: it's responsible for creating verification keys ("scriptPubKeys" in Bitcoin, which are usually abbreviated to addresses) whose corresponding private keys are controlled by the correct parties to the correct extent, and for correctly and securely storing these keys. Wallets are decoupled from the main cryptocurrency cryptosystem, and in particular are not part of the
hard part --- consensus code --- but they are certainly cryptographic and are subject to the same sort of subtle missteps as other cryptosystems. (For example, the oft-cited BIP32 "bug" where a party in possession of a public key and chaincode can derive the secret key from a non-hardened child secret key; it would not be hard to come up with a plausible-sounding system which "unintentionally" exposed secret data through this mechanism.)
I'll repeat my above statement: basically the entirety of cryptocurrency really
is cryptography. This is important. It's why things are so subtle, why complexity is dangerous, and why changing even trivial-seeming things can have drastic and hard-to-analyze consequences. This is where "a lot of harmful garbage and dysfunctional software" comes from. It's the default for poorly-thought-out systems. And in cryptography, "poorly thought out" means anything less than expert cryptographers spending large amounts of time and effort designing things to be both correct and
clearly correct. (Given how few experts there are in the cryptocurrency space, I could tell you that almost all of it is shit just by the pigeonhole principle
.)
This is not a cheap standard to hold a system to even if its designers want to; and falsely claiming that something is not cryptographic is an easy way to excuse not wanting to. But such claims do not change reality.
The encryption of data through cryptography is the foundation of cryptocurrencies.
This is simply false.
Hi Andy,
Sure, I understand cryptography is key for cryptocurrencies. I was just proposing some semantic misunderstanding of what Tok said because he was getting it in the neck
V. interesting reading this insight into cryptography, thanks.
I get the value, it's the building block of privacy and freedom in the digital age, and in cryptocurrencies it enables all the good stuff like proof of ownership, trustless transfer, global uniqueness etc as well as the low level workings. And humans are useless for cryptographic function because our actions are non-deterministic...thankfully. I'm into etymology so - information, literally 'the ability to shape people's actions' ergo 'information is power'. Cryptography aka 'hidden writing' or ability to (as i now understand) 'confine and restrict' the ability to shape people's actions...I think that's the important bit for freedom in the digital age especially when governments can physically access our info but we can't access theirs.
I see what you mean about masternodes but I think it's arguably not a purely cryptographic function because from what you are saying the main purpose of cryptographics is to confine constrain information as gmaxwell says for the purposes of un-confining disseminating it later (as I understand it as a layperson). Masternodes don't want to confine the information they want to 'lose' it best as possible and they do that by adding identical information from others at the same time to create ambiguity which is getting more into information theory and entropy than cryptography right? (if I understood correctly, although there is still arguable confinement and containment...). but semantics aside I don't think anonymity has to be a cryptographic process, e.g. for me i don't want a reversible process just want the info as lost as possible (Analogy would be people sitting around a table mixing their $ bills together: Cryptographic approach would be sequentially logging what went were and then confining constraining that log information using a cipher or whatever. Masternode approach would be blindfolding them first and just randomly mixing them, maybe keep a log and randomize it, IDK enough about the inner workings..). Cryptography seems like best fit for XMR anonimity because they want the reversibility for viewkeys etc but is cryptography best fit for trying to 'lose' information as drk/Dash is aiming to do?
If that's correct then XMR DRK are really different and it depends what the cryptocurrency users are looking for, DRK anonymity is designed to (ideally) 'lose' transaction information at a given moment through adding ambiguity, XMR anonymity is designed to 'confine and constrain' transaction information using a cryptographic function but provide the ability to decode the information (ideally) only by those who are authorized by the encoder.
The threat models of compromising the anonymity for each system would then be, for DRK, to acquire enough information to circumvent the ambiguity process meaning high control of the masternode network, or for XMR, to compromise the cryptographic functions governing the ring signatures, or the system that implements those, or extort the 'viewkey' (my XMR knowledge is limited).
So IDK...I am just a drk miner so I don't speak for anyone in DRK (and I dropped out after Calculus so i could have easily interpreted it wrong..)
BTW...I never spoke with a cryptographer, what you think about number stations?
cheers
