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

The Verge Hack, Explained

blog.theabacus.io

Post reply on HN
Print view View original

41–50 of 68 posts

Re: The Verge Hack, Explained

#41

Post by danbruc »

EGreg wrote: post #33
danbruc wrote: post #19

Earlier quoted context omitted.

You are mixing two things up - figuring out the time an event happened and trusting a timestamp that someone attached to an event. Within a single inertial reference frame there is a canonical way to order events, namely by the time coordinate in that reference frame which is the same everywhere. And for all practical purposes of building a distributed system on Earth, Earth is an internal reference frame. Ensuring t…

You didn’t explain how a distributed byzantine system can synchronize a global time that’s continuous. I am saying that you won’t be able to — local time is continuous and you should use that for comparisons. It’s nowhere near as straightforward as you think to make global time work. You will find that “making sure nobody can or will cheat” will require exactly the kind of effort that will make the concept of a globa…

I only object your claim that relativity - as in special or general theory of relativity - has any bearing on the design of distributed systems on Earth. In particular, you can order events occurring at different places if you have properly synchronized clocks. The fact that a common solution is to use logical clocks like Lamport clocks is because it is easier to do so than to maintain a network of precisely synchronized clocks. But that does not mean it can not be done with synchronized clocks, see for example Google's Spanner [1] database.

I never addressed what else is necessary to make this work in the presence of Byzantine faults. But I am also pretty sure that just sending information about events from A and B to C and asking C about the ordering is not sufficient, after all C may be faulty and give a wrong answer, change its mind after every query, or just not respond at all. Admittedly you did not provide many details, so I may very well totally misunderstand what you had in mind.

But, just to repeat, I am not really addressing the requirements of a Byzantine fault tolerant system, I am just saying that the special or general theory of relativity is not relevant for distributed systems on Earth.

[1] https://en.wikipedia.org/wiki/Spanner_(database)

Re: The Verge Hack, Explained

#43

Post by EGreg »

danbruc wrote: post #41
EGreg wrote: post #33

Earlier quoted context omitted.

You didn’t explain how a distributed byzantine system can synchronize a global time that’s continuous. I am saying that you won’t be able to — local time is continuous and you should use that for comparisons. It’s nowhere near as straightforward as you think to make global time work. You will find that “making sure nobody can or will cheat” will require exactly the kind of effort that will make the concept of a globa…

I only object your claim that relativity - as in special or general theory of relativity - has any bearing on the design of distributed systems on Earth. In particular, you can order events occurring at different places if you have properly synchronized clocks. The fact that a common solution is to use logical clocks like Lamport clocks is because it is easier to do so than to maintain a network of precisely synchron…

I think you misunderstood my point.

I said that relativity is a special case of this principle that time is a local phenomenon.

Re: The Verge Hack, Explained

#44

Post by rando444 »

I'm very curious about the price increase.

The only thing that comes to mind how something like that would be possible is if the news trends generated by this were somehow driving algorithmic trading .. either that or people investing in something they're just hearing of for the first time.. But even these possibilities don't seem rational.

It just doesn't seem to make much sense... unless the whole thing was being driven and pushed by some larger unknown actor with a lot of resources .. (state government, goldman sachs, etc.)

Re: The Verge Hack, Explained

#45

Post by XR0CSWV3h3kZWg »

pitaj wrote: post #32
EGreg wrote: post #5

Earlier quoted context omitted.

PoW is leader based consensus. It’s slow and can’t handle millions of simultaneous payments. Why do new distributed ledger technologies still prefer to use it, when there are other, far faster consensus algorithms out there?

BCH at 32 MB blocks can handle 108 Tx/s. It would only need to grow to 6 GB blocks to handle the 20k Tx/s that VISA claims they can handle. It would only need to grow to 600 MB blocks to handle the 2k Tx/s that VISA handles on average. 600MB every 10 minutes is no sweat for even modest internet connections nowadays. With technology like weak blocks and the lighting network and other side-chains the network can handle…

This is pretty offtopic, but the amount of time it takes to download the block isn't the only factor, amount of time to process & verify the block is pretty important as well as the storage costs etc.

That's all pretty moot at the moment as BCH is averaging ~50kb blocks.

Re: The Verge Hack, Explained

#46

Post by sterlind »

This hack underscores the need for formal modeling systems like TLA+ or Coq. Essentially, you define successor state axioms, where:

1. start with a seed state.

2. non-deterministically choose the next step (e.g. submit a tx)

3. ask the system to prove the inverse of an invariant you want to check (e.g. hash power cannot fall below X% of total network power.)

Decentralized protocols have bugs; trust only what you can prove.

Re: The Verge Hack, Explained

#48

Post by sean2 »

TheDong wrote: post #31
danielvf wrote: post #13

"Exactly how the largest protocol-level hack of a cryptocurrency in recent memory could preceed said cryptocurrency increasing in price and then announcing a partnership with the most trafficked porn-site on the internet is a question I’m forced to leave open-ended; my personal pet-theory is that it has something to do with the fact that the world makes no sense and human beings are all completely out of their goddam…

Please don't quote portions of the article with no added commentary or insight. We're all capable of reading and recognizing well-written portions of the posted content.

This was a particularly interesting bit though and I was hoping some people might explain/comment more about it.

It seems that getting your crypto-currency hacked gives publicity and boosts value! See also: the Experian hack.

Re: The Verge Hack, Explained

#49

Post by danbruc »

EGreg wrote: post #43
danbruc wrote: post #41

Earlier quoted context omitted.

I only object your claim that relativity - as in special or general theory of relativity - has any bearing on the design of distributed systems on Earth. In particular, you can order events occurring at different places if you have properly synchronized clocks. The fact that a common solution is to use logical clocks like Lamport clocks is because it is easier to do so than to maintain a network of precisely synchron…

I think you misunderstood my point. I said that relativity is a special case of this principle that time is a local phenomenon.

Which is not really true, time is frame depended, not location depended. Just take the twin paradox, when they start and later meet again, they are in the same place but different amounts of time have passed for each of them. The closest thing I can think of is that the ordering of two events is not ambiguous if they are timelike separated while for spacelike separated events you can find different reference frames which disagree on the order of the events.

Re: The Verge Hack, Explained

#50

Post by EGreg »

danbruc wrote: post #49
EGreg wrote: post #43

Earlier quoted context omitted.

I think you misunderstood my point. I said that relativity is a special case of this principle that time is a local phenomenon.

Which is not really true, time is frame depended, not location depended. Just take the twin paradox, when they start and later meet again, they are in the same place but different amounts of time have passed for each of them. The closest thing I can think of is that the ordering of two events is not ambiguous if they are timelike separated while for spacelike separated events you can find different reference frames w…

What you said doesn’t disprove what I said :)
Post reply on HN