Want to take part in these discussions? Sign in if you have an account, or apply for one below
Vanilla 1.1.10 is a product of Lussumo. More Information: Documentation, Community Support.
I have a few thoughts on how one could tweak the notion of proof of work to avoid use of massive amounts of computational resources; I can try to write something about that here later.
Proof of work is past. Modern blockchains use proof of stake or other algorithms. Look at Durov’s TON paper.
Blockchain is invented in Wei’s 1998 paper. Nakamoto rediscovered it in 2008.
Ideally, in future public blockchain products lattice cryptography should be used to ensure staying well after quantum computing becomes reality.
Hi Zoran, please feel free to update the references :-).
Proof of work is past. Modern blockchains use proof of stake or other algorithms. Look at Durov’s TON paper.
I am not an expert, but I am far from sure that this is the case, neither in practise nor theoretically. There are numerous problems with proof of stake as well, and I have not seen any alternative that is as secure as proof of work. This is precisely because proof of work, in its crudeness, relies on something that is impossible to get around without expenditure, i.e. actual computational power (and attendent electricity, etc).
As I say, I do have some ideas on how to tweak proof of work to avoid the resource burning. But if you can convince me that some version of proof of stake is already superior, I would be very happy to hear your arguments :-).
I will elaborate on this in near future. Surely, I agree that PoS algorithms are not solving all the balance problems in blickchajn ecosystems. But in most cases it is not the security which is the critical loss when passing to proof of stake. The fact that in addition to straightforward mining there are also useful and unavoidable comoutations, like confirming the validity of transactions, smart contract execution and alike, one can measure contributions of those performing these community service operations and allow those contributors, in respective proportions, to ALLOW to mine. In other words, you set mining problems computationally easier to solve (hence avoiding ASIC competition, but still nontrivial to solve) but you allow only some amount of new mined tokens to those authorized. On the other hand, even being one of validator nodes, is also preselected using say currency stake, rotating orders anx consensus.
I expanded the entry a bit and crosslinked it, including with my own bibliography entry and new stub smart contract.
I have added about a dozen of key papers in blockchain. I also created a related page Vitalik Buterin and added few items at Nikolai Durov.
Extended literature lists are kept in my personal lab, blockchain (zoranskoda), smart contract (zoranskoda), EOS (zoranskoda), distributed consensus (zoranskoda), distributed computing (zoranskoda), Telegram Open Network (zoranskoda), Ethereum (zoranskoda), bitcoin (zoranskoda), payment blockchain (zoranskoda), economics of blockchain (zoranskoda), blockchain security (zoranskoda), high performance DLs (zoranskoda), fintech (zoranskoda) etc.
By the way, looking over this entry, the section Criticism (here) needs attention.
Its first paragraph is out of place for the Lab (as much as it would be, say, for the entry LHC to debate CERN’s electricity bill). Let us stick to math, physics and philosophy and leave connecting to the latest zeitgeist fads to the reader who feels the need.
Its second paragraph touches on the noteworthy technical point of Byzantine Fault Tolerance, without however saying so. This would be a technical point worth expanding on.
While it was not written by me, I think I would rather support leaving mentioning the energy as a factor which does concern the algorithmic security. Namely, the blockchain security is mostly about incentivization — if having enough resources one can break it. Now, the block reward for PoW is not the only place which costs and where miners can earn. There is for example fee for transactions and for performing smart contracts. Now, in bitcoin for example, in mid 21 century the block reward will diminish and disappear eventually and all incentive will be from transaction and other fees. That means that the security incentive will be small and that, according to Buterin’s recent statement, the security of bitcoin network will drop by orders of magnitude. So, understanding all that economy is very much related to the calculation of the security factors which are of algorithmic importance.
P.S. the question of energy is mentioned under criticism for bitcoin/PoW systems in wikipedia where however the main criticism from the blockchain community itself is not mentioned at all. Namely, by being first and most primitive and slow, blockchain has large support from the public which draws huge Ponzi scheme-mechanisms of investing into bitcoin, and as a consequence, much more useful and advanced blockchain systems with high transaction speeds and advanced smart contracting built in are not getting investor’s attention. Thus the technical advancement of blockchain is slowed down by the “blockchain dominance”. Many hope(d) for so called “inversion” where the alternative projects would get more funding and their cryptos would get bigger market share relative to bitcoin.
Thanks. That’s all interesting, but it’s about the practice of cryptocurrencies (which is marginally on-topic here) not about the concept of blockchains as such. Compare to entries such as say, nuclear force, where it would be off-topic to debate politics of power plant installation.
Let’s have an entry cryptocurrency, if we must, to discuss such matters there, and try to steer the entry blockchain to be a sober reference in the area of computer science.
I have renamed the (tiny) section “Criticism” to “Fault tolerance”, reworded to make explicit the notion of “Byzantine Fault Tolerance” (BFT) with pointer to:
(just the first best reference I could find.)
and deleted the sentence making unreferenced claims about “sustainability” (which is for another entry, maybe cryptocurrency, or maybe even for another Wiki).
(This reminds me of an anecdote: As a kid I bought a book on Amiga PCs which explained some file format with a comment along the lines that: “A field of 4 Bytes is more than sufficient for recording the file length, since nobody will ever fill more than bytes with data.”)
Just for the record. The entry makes an impression that it is essential for blockchain to be blockchain to use the proof of work, while this is just one of several possible protocols, nowdays variants of PoS (proof of stake) protocol being the dominant framework.
12: I am not very enthusiastic about creating a page on “cryptocurrency”, but if it were created let me state an important distinction: the general public (and the exchanges) nowdays calls any valueable token on blockchain a cryptocurrency. Traditionally, however, only THE token which figures its role in the incentivization scheme of the base protocol of the blockchain is the cryptocurrency. Thus one (traditional) blockchain has at most one cryptocurrency and possibly many kinds of tokens. Everything else are just the tokens which happen to be recorded somewhere in the system, not much different from reward “miles” from the air companies.
I am not going to create the entry, but feel free to do so. Let’s just try to keep it on questions of a vaguely ontopic (computer-)scientific nature and not go down the numerous rabbit holes waiting in the territory. (Happy to explore them, but not on the nLab…)
For instance, I find it interesting to see increasing interest in type-theoretic software certification used in the field, as witnessed by the references that I had added the other day (here). If we could have more on that angle, I’d be very interested.
1 to 16 of 16