Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Blockchain at Berkeley

blockchain.berkeley.edu

Post reply on HN
Print view View original

41–50 of 63 posts

Re: Blockchain at Berkeley

#41

Post by DennisP »

delhanty wrote: post #21
tylersmith wrote: post #10

Earlier quoted context omitted.

It's more of a tree with single root and a function returning the longest tip by some distance method. Bitcoin uses amount of work proved as the distance. Bitcoin ignores other tips unless/until one of them becomes longer than the current tip but some systems like Ethereum leverage stagnant branches for their length function. But yes, dynamic-membership multi-party signature schemes and their use for creating permiss…

Straight question for the uninformed (me): Do you (or anyone else) have a technical reference for differences between the Bitcoin and Ethereum blockchain length function?

Bitcoin's method is detailed in Satoshi's classic paper:

https://bitcoin.org/bitcoin.pdf

Ethereum uses a variant of the GHOST algorithm, which was introduced in the paper Accelerating Bitcoin’s Transaction Processing: Fast Money Grows on Trees, Not Chains.

https://eprint.iacr.org/2013/881.pdf

The Ethereum white paper mentions how they modified it:

https://github.com/ethereum/wiki/wiki/White-Paper#modified-g...

Full detail is in their infamously difficult Yellow Paper:

https://github.com/ethereum/yellowpaper

Re: Blockchain at Berkeley

#42

Post by Klathmon »

akhosla wrote: post #36
xg15 wrote: post #29

Earlier quoted context omitted.

If you talk about the general idea of an append-only distributed database with no central point of control, I agree with you. However the proof-of-work/auto-adjusting-difficulty method of asserting truth still seems horrible to me from a technical perspective. Politically, I don't want to live in a world where this principle catches on and an ever-growing % of energy production would be spend on solving meaningless m…

Recommend you check out Proof-of-Stake. It solves the problem of power wastage. ;)

But many people (including those who are actively developing it) aren't even sure if Proof-of-Stake is even fully possible.

As far as I know there are still some pretty big holes in PoS that still need to be filled in.

Re: Blockchain at Berkeley

#43

Post by wslh »

arthurcolle wrote: post #5

'Blockchain' is literally the least interesting part of cryptocurrencies and their applications. its a boring flat file structure that is an artifact of needing many clients independently deriving a shared ledger, nothing more and nothing less. the interesting part is how you can bootstrap issuance of a new asset by incentivizing participants to assist in maintaining a specified computer network in exchange for crypt…

Sorry, but I don't think you understand what a public blockchain is and the improvements/innovations that it requires to sustain a higher volume with faster blocks creation. This is connected to what you said about issuing tokens: last week Status.im (shameless plug: security audited by my company) related transactions caused a DDoS over the Ethereum network, which it means that we need to improve the current blockchains to support the kind of innovations you like.

On the other hand, lowering the barrier for issuing equity or app coins make creating scams or tempting people to run with a lot of money trivial. A recent Fred Wilson's post[1] about this makes the point very clear.

Lastly, not everything can be descentralized because you are challenging computation limitations and/or you need oracles outside the network that are the weakest point of the whole system.

[1] Buyer Beware - VC Fred Wilson on ICOs https://www.reddit.com/r/ethereum/comments/6h7d0j/buyer_bewa...

Re: Blockchain at Berkeley

#44

Post by delhanty »

DennisP wrote: post #41
delhanty wrote: post #21

Earlier quoted context omitted.

Straight question for the uninformed (me): Do you (or anyone else) have a technical reference for differences between the Bitcoin and Ethereum blockchain length function?

Bitcoin's method is detailed in Satoshi's classic paper: https://bitcoin.org/bitcoin.pdf Ethereum uses a variant of the GHOST algorithm, which was introduced in the paper Accelerating Bitcoin’s Transaction Processing: Fast Money Grows on Trees, Not Chains . https://eprint.iacr.org/2013/881.pdf The Ethereum white paper mentions how they modified it: https://github.com/ethereum/wiki/wiki/White-Paper#modified-g... Full…

Thank you for all those links!

I had some idea about the Bitcoin blockchain - also succinctly summarized by @tylerswift.

But the Ethereum material is all new to me. It at least sounds more interesting than "the least interesting".

Re: Blockchain at Berkeley

#45

Post by icebraining »

kobeya wrote: post #18
avaer wrote: post #7

Earlier quoted context omitted.

There's a bit more to it. Proof of work hashing and ledger consensus rules and difficulty targets and fee schedules and forks. It's still all standard bachelor's degree algorithms, and I agree with the gp that it's not the interesting part. The interesting part is using these algorithms for social consensus on value.

I disagree those are essential. Federated chains like Elements Alpha are also considered block chains but lack many of those properties. I think it really is the case that a "block chain" is merely a singly linked list with hash linkages.

But that's just an hash chain, it has existed for longer than the blockchain. For example, all digital invoices in my country are hash-chained, to prevent the merchant from swapping one for another, and this standard was published before Bitcoin.

Re: Blockchain at Berkeley

#46

Post by wfunction »

abattleroyale wrote: post #17
vostrocity wrote: post #14

Earlier quoted context omitted.

Knowing student groups, it's not about making money, but a few smart undergrads showing other undergrads emerging fields and letting them get some hands-on experience. A group like this has a DeCal (a student-led university-credit course), events, speakers, and probably a selected group that works on club leadership and consulting.

It's definitely a mix of the two, speaking from experience.

Experience with this club, or experience with some other random club?

Re: Blockchain at Berkeley

#48

Post by DennisP »

icebraining wrote: post #45
kobeya wrote: post #18

Earlier quoted context omitted.

I disagree those are essential. Federated chains like Elements Alpha are also considered block chains but lack many of those properties. I think it really is the case that a "block chain" is merely a singly linked list with hash linkages.

But that's just an hash chain, it has existed for longer than the blockchain. For example, all digital invoices in my country are hash-chained, to prevent the merchant from swapping one for another, and this standard was published before Bitcoin.

That sounds interesting. What is the standard called? I'd love to read more about it.

Re: Blockchain at Berkeley

#49

Post by rocqua »

akhosla wrote: post #36
xg15 wrote: post #29

Earlier quoted context omitted.

If you talk about the general idea of an append-only distributed database with no central point of control, I agree with you. However the proof-of-work/auto-adjusting-difficulty method of asserting truth still seems horrible to me from a technical perspective. Politically, I don't want to live in a world where this principle catches on and an ever-growing % of energy production would be spend on solving meaningless m…

Recommend you check out Proof-of-Stake. It solves the problem of power wastage. ;)

That requires solving the 'nothing at stake' problem. I recall reading some proposed solutions, but they were complicated.

If anything is great about proof-of-work, it's how simple it is.

Re: Blockchain at Berkeley

#50

Post by throwaway5752 »

arthurcolle wrote: post #5

'Blockchain' is literally the least interesting part of cryptocurrencies and their applications. its a boring flat file structure that is an artifact of needing many clients independently deriving a shared ledger, nothing more and nothing less. the interesting part is how you can bootstrap issuance of a new asset by incentivizing participants to assist in maintaining a specified computer network in exchange for crypt…

Well, that is your opinion. I personally think that all public cryptocurrencies are eventually worth precisely zero because of the of the nature of probabilistic data loss due to hardware failure or otherwise, and the only thing holding them up is the honeymoon phase of the mining gold rushes. Blockchain (and associated distributed consensus systems) is interesting, though, for no-trust distributed workflows.
Post reply on HN