Topic: Gold collapsing. Bitcoin UP. - page 585. (Read 2032291 times)

I did read it, and also the "sidechains for dummies" blogpost that people recommended, and some more posts.

So what you are saying is that I did not understand it at all.  Perhaps.  But then I ask, for example, why Bitstamp is not already a "sidechain".  OK, it is not decentralized nor merge-mined, but does the whitepaper say that a sidechain must be those things?

My impression is that the paper did not want to rule out anything, for fear that it might prevent co-opting a possible "bitcoin killer".

So I take it you didn't read the paper 

Satoshi's whitpaper is not a "scientific" paper but a good "technology advance" paper.  It specifies a concrete real-world problem (build an e-cash system that does not depend on trusted 3rd parties) and an open sub-problem of that (how to motivate the volunteers to maintain and secure the ledger).  It then describes a proposed solution for that sub-problem, with enough detail that any competent programmer could implement it.  The paper concludes with some logical arguments and computations showing that the proposed solution does not have some flaws that one may think of.  Then Satoshi built an implementation that would be sufficient to prove the soundness of the solution.  He set that implementation running, and recruited enough interested people to make the test meaningful.

The Sidechains whitepaper is nothing like that.  It does not define concrete real-world problem(s) to be solved, and then develop suitable solution(s) for it/them.  Instead, the "problem" that it sets to solve is how to save bitcoin from being superseded by centralized payment systems and other altcoins.  The "solution" they propose is basically to rebrand the best  altcoins plus some unspecified array of other network services as "sidechains of bitcoin",  in exchange for some unspecified advantages involving merged mining and the prestige of bitcoin.  However, there are no details in the paper to even tell what applications could be "sidechains", how the integration would work, what concrete benefits they would derive from that, why other projects could not obtain the same benefits without being "sidechains of bitcoin" -- and why would bitcoin be saved as a result.  Needless to say, there is also no analysis of possible failure modes.

Just as bitcoin reminds me of the Wright Brothers' Flyer One, the sidechains proposal reminds me of their unsuccessful efforts to commercialize their invention.  (AFAIK, the first commercially successful series-produced airplane, still only a toy for aeronautics nerds, was this one.)

"Dear investors, you are all aware of the hard times that Ford Motor Co. has been going through, because of competition from manufacturers of cars that are cheaper, faster, more beautiful, or more reliable than ours.  You know that, while we could incorporate their advantages in our Model T, modifying our assembly lines is very expensive and would take too long.  But we have come up with an idea that will surely save us.  We will allow the other manufacturers to rebrand themselves: "Ford Jeep" instead of "Jeep", "Ford Jaguar" instead of "Jaguar", "Ford Mercedes" instead of "Mercedes", etc..  We will allow them to use our great assembly lines, as long as they don't require any special change or interfere with our production.  We will also build a yard where our customers will be able to safely trade their Model Ts for cars of our competitors, at values established by the latter.  Thus we expect to turn our most formidable competitors into our allies, and, with their help, retain our dominant share of the billion-dollar revenue to be made in the car market."

Anyone who can manage to get kicked out of trolltalk is worth listening to in my book.  For instance, Atlas was the ass-clown's ass-clown but he actually produced some very insightful peals if one looked closely enough.  MP (of his sock-puppet pimpette thing or whatever it was) was wildly amusing to me in general, and often enough cast pearls.

And I am natively closer to that view of blocksize anyway of course.

edit - spellchecker: naively --> natively (is that a word?)  Ya, ya, Freudian slip.  yuck-yuck.

Early on I saw lots of opportunity for business services for expand transactions, many business ideas will be enhanced by a limited block size.

I think the idea enables an alternate outcome I now feel the primary experiment is unlimited block size and a secondary more radical experiment is limiting the number of transactions.

The reason being storing the blockchain is priced in to the incentive structure for those who have an invested interest in its integrity, it is a service the wealthy provide to the network for using the network.

My opinion is that in terms of both science and engineering, the software industry is kind of a joke, but I've never had any involvement in the academic side of things.  That lack of rigor may explain why shit actually gets done...

In a practical sense, it's damn near impossible for something like these non-academic crypto whitepapers or ideas to obtain anything remotely resembling quality 'peer review'.  It's to much work for to little gain, and the signal to noise ratio in crypto-land (aka, 'spam') is a giant factor.  There are relatively few people who are even qualified to do the kinds of review necessary and they have more interesting things to do.  Poul-Henning Kamp's struggles to get someone to review his full disk encryption work is an interesting read which might help understand the problem(s).  I don't have a link handy.

Bitcoin seems to owe most of it's success to Hal taking a modest interest at the right time.  A splash of diesel at the right time can mean the difference between a fully engulfed fire pile or one with not a wisp of smoke a few hours later.  (I happen to have been working on some brush clearing over the last week and stop by to check my favorite thread when I need a break.)

The stuff from that Israeli dude (presented at the SJ 2013 con) which seems to be becoming the snark ideas is one exception to this lack of rigor (in my read of things.)  It's just so damn interesting and universally important that it can get the kind of attention needed.

My other current favorite thread here is the one about global warming.  I finally took the dive into researching that aspect of 'our common future' recently after blowing it off for years.  My respect for the process of academic and scientific rigor has been quashed as effectively as my aspirations for 'Socialism' as a result of this research.  Crypto-currency may be better off without 'peer review' in the end.

+1

In fact one's knowledge and understanding grows where one put there focus. Many have been enabled by Bitcoin's rapid price assent and have contemplated daily how this technology will propagate and succeed, this economic thought experiment has been undertaken by a fiew who spend there days assimilating the communities synaptic sparks.  Many developers spend more time developing than they spend conceiving the concepts they develop they don't have the luxury of spending 4 years in thought about how human behaviour will react to the logic before they start developing.

Some even get hung up on there lack of understanding and have sleepless nights and urgently set out to change things they feel they know well.

I'm curious to know what you (and others btw) make of Mircea Popescu and his crew's position that blocksize should not (arguably never) be increased

i have a question for the computer scientists in the crowd. 

i've noticed some things about the WP's over the last coupla years, including Satoshi's, that i wasn't aware of. first, there doesn't seem to be any peer review.  not only that but what strikes me is that it seems that authors come up with a particular logic pattern that makes sense to them and then construct maths to support that logic.  i often don't even see large data sets, testing, or simulations performed.  that supposedly is a standard for proof of a theory.  is this correct?  this is not a criticism of the CS field but just an inquiry to help me understand your field.

in many fields, a hypothesis is asked, then an experiment with blinded investigators along with large #'s of subjects and controls is carried out over a statistically significant time period to rule out investigator bias (can never do that completely) and prove that the hypothesis is correct within a std dev often of p<0.05 or 5%.  i realize these are different methods in different fields so it's hard to determine which is more rigorous.

much of what drives Bitcoin is logic and the prediction of human behavior.

geeks don't have a monopoly on that.

this headers only first download capability just got implemented with the last update.  i thought you said we were already at the stodgy stage?

Not to be offensive, but most of what you say about most things (economic, engineering, technology, investment, etc) is very primitive and often enough laughable incorrect.  Another short-hand term I've used in the recent past to describe this is 'ass-clown'.