Author

Topic: Important Topics for Review Articles (Read 817 times)

sr. member
Activity: 333
Merit: 252
May 15, 2015, 02:30:23 AM
#3
it seems to me that on this topic it is still original research that is missing, rather than reviews.
That is, what is missing is a formalization of  what it means to achieve a provably secure trustless consensus, and proving that PoW/ PoS/other alternatives achieve/ do not achieve it.  Of course, there may be several stronger/weaker/incomparable notions of such "provably secure consensus", whose merits  and practical applicability could then be argued.

Maybe a proper review of the existing literature is a good start though.
legendary
Activity: 1162
Merit: 1007
May 12, 2015, 10:38:19 PM
#2
I think one very useful article could review the history of electronic cash, focusing on the problem of trustless consensus.  It could formalize the consensus problem that Bitcoin's PoW solves, and then compare and contrast that to other proposed solutions such as proof-of-stake and the ripple/stellar protocol.

I know efforts have been made to do this (see below), but I think there's still a sort of "impedance mismatch" between the content that's been generated and the audience who would benefit from being exposed to it.  For example, many newbies are drawn to PoS like bugs to a porch light; I think these newbies see the PoS/PoW debate as more a "matter of opinion" than a matter of what is actually known to work versus what isn't.  Even high-level thinkers like David Andolfatto  (VP Federal Reserve Bank of St Louis) fail to comprehend the breakthrough that Satoshi made with the PoW security model:

Quote from: David Andolfatto (VP Federal Reserve Bank of St Louis)
Is there a cheaper way? The current protocol uses what is called "Proof of Work" and there is very much something like a common resource problem here, with "overinvestment" in computing power (relative to first-best). But the community is working on alternatives, like Proof-of-Stake. But at present, there appear to be trade-offs. Cheaper protocols are also less secure, at least for now. But I expect a big break through in the not to distant future.

From: http://andolfatto.blogspot.ca/2014/11/bitcoiners-surely-we-can-do-buiter-than.html?showComment=1417191644448

A well-written "review article" on this topic published in a peer-reviewed journal would do a few things:
    1. It would convincingly show the extent of our knowledge related to PoW (i.e., that it's provably secure under certain assumptions) as well as our lack of knowledge related to the security of other consensus mechanisms (PoS/ripple/etc).
    2. It would give the work a stamp of authority so that references to the article (in forum posts, conversations, or when giving talks) would carry more credibility.  (This "stamp of authority" is especially useful for convincing people who don't possess the technical background need to evaluate the actual arguments).  
    3. It would raise the bar for proponents of other consensus mechanisms to formalize the security assumption of their proposed cryptosystems. (E.g., "So you're new variant of PoS solves problems X, Y, and Z? Why hasn't it been it been formalized and published in a peer-reviewed journal like the Satoshi PoW solution has?)

Notable efforts have already been made.  On the first page of the Sidechains white paper, the authors give a good review of e-cash starting with Chaum's work from 1983, motivating the need for an anonymous/dynamic membership approach to avoid Chaum's trusted clearing house.  They describe Adam Back's Hashcash proof-of-work as a method to avoid the Sybil attack problem in anonymous networks, and then discuss how Satoshi put the right pieces together in a way that provides an economic incentives for anonymous participants to act cooperatively (thereby allowing anonymous membership without risk of a Sybil attack).  Lastly, they introduce the concept of a dynamic-membership multi-party signature (or DMMS) and describe Bitcoin's blockheaders as examples of such signatures.  

In other work, Andrew Poelstra contrasts Bitcoin's PoW system to PoS systems in Distributed Consensus from Proof of Stake is Impossible and the newer version "On Stake and Consensus". I really like his interpretation that mining translates a "fact of physics to a fact of mathematics" and I appreciate the accessible language that the paper is written in.  However, I'm not sure this piece formalizes the problem to a sufficiently-convincing extent.

Andrew Miller and Joseph J. LaViola, Jr. work towards formalizing the trustless consensus problem here, and in perilous mathematically detail, concluding that Bitcoin's PoW mining "achieves consensus...except for a negligible probability, when Byzantine faults make up less than half the network."  I'd like to see a more digestible version of these results, especially if it's possible to formalize PoS in a similar way (and outline the differences, for example, that PoS relies on what Vitalik et al. refer to as "weak subjectivity.")

DeathAndTaxes makes several insightful comments starting here as well as in many other threads.  

TL/DR: To answer my own question, I'd like to see a review paper that sort of fuses all of the above work together.  
legendary
Activity: 1162
Merit: 1007
April 22, 2015, 07:13:23 PM
#1
When I discovered Bitcoin in early 2013, I was blown away by the brilliant ideas and content generated by this largely informal and online community.  At the same time, as a new member trying to learn about Bitcoin, it was difficult for me to distinguish the "signal from the noise" when the media consisted of forum posts, blogs, and self-published papers.  What I wanted was an authoritative way to identify high-quality cryptocurrency research that was generally accepted as accurate and important by experts in the field.  

Something that I think would help increase the signal-to-noise ratio is efforts aimed at aggregating and filtering this content.  Basically, seeking out the relevant work generated across both formal (journal papers, conference proceedings) and informal channels (forums, reddit, github, etc) and packaging it into neat and tidy review articles with thorough referencing.  The purpose of this thread is to discuss what topics related to Bitcoin would be best served by such review articles...

...So, if you could press a button and instantly the community would have access to a well-written/researched peer-reviewed "review article," what topic (or topics) would you pick?

Jump to: