8 ms·

For those interested in understanding the tech rather than the typical bashing things as beneath them, I wrote up a detailed technical explainer of how Ethereum

For those interested in understanding the tech rather than the typical bashing things as beneath them, I wrote up a detailed technical explainer of how Ethereum PoS works: https://0xfoobar.substack.com/p/ethereum-proof-of-stake https://0xfoobar.substack.com/p/ethereum-proof-of-stake

SilasX 4y ago
Guess you can update this part now! >3. PoS (Ethereum, soon™)
- 0xfoobar 4y ago
  Indeed! :)
atomlib 4y ago
What's Substack? Is that new Medium?
- speedylight 4y ago
  That’s how I always thought about it yeah.
- NavinF 4y ago
  Yeah it's like Medium except it has less fluff/self-promotion.
- lloppal 4y ago
  It's like Medium but wants you to pay for newsletters.
  - polisteps 4y ago
    Which is fairer imo than paying for the whole site. I prefer creator control on which content is paid and which content is not.
- toastal 4y ago
  It'll be okay like Medium used to be until it isn't and the funders turn it into Medium while seeking that return on investment. You're still better off owning your own blog in the long run.
bambax 4y ago
Isn't it a little unfortunate that the acronym for Proof of Stake would be PoS? Couldn't they think of something else?
- kwhitefoot 4y ago
  You mean people might confuse it with Point of Sale? /s
  - capableweb 4y ago
    Don't be obtuse. They clearly imply people might think it means Point of Service.
  - butterlesstoast 4y ago
    I don’t think that’s an obtuse comment. I understood the article to be talking about Point of Sale the entire time because I live off the blockchain hype.
    - Consultant32452 4y ago
      Hopefully the new algorithm is not a Piece of Shit.
    - kwhitefoot 4y ago
      I think you missed the joke. Or was it? Hmm, now I'm wondering!
- kortex 4y ago
  I'm sure you can find a naughty word or phrase, or some other semantic collision, out of almost any 3-letter combination, in some language somewhere. Sure it's one of the more well known ones in a very common lingua franca, but we humans are smart and can context switch. Edit: Actually not sure if you are talking about Point of Sale, Place of Service, Piece of $#!7, or something else. Case in point: context matters.
bsaul 4y ago
I find the complexity of that algorithm both impressive ( since they seem to have made it work), but also quite worrying. I'm really not sure how such a beast can't be filled with bugs, not in the implementation but rather in the protocol. I know a lot of very smart people are working on this, but i'd rather have something conceptually simpler to work as the base layer for a whole new economy.
- thrown_22 4y ago
- Hendrikto 4y ago
  Making PoS scale to hundreds of thousands of nodes with commodity hardware is not simple though. Few projects managed so far, and Ethereum wasn‘t designed for it from the outset, so it‘s even more difficult.
  - preseinger 4y ago
    But of course no system can scale to O(100k) nodes and retain reasonable availability and consistency properties.
  - valzam 4y ago
    Of course they aren't really scaling to hundreds of thousands of nodes. At any given time only 120 nodes are involved in consensus.
- arlcode 4y ago
  Cryptocurrencies impose some complex constraints on themselves that require complex solutions. Conceptually banks and exchanges solved the consensus problem decades ago and they did it with a highly secured simple database and lots of crosschecks. But if you trust nobody (except some developers somehwere) then things get tricky
  - chompychop 4y ago
    Could you go into more detail or provide references on where I could read up more on how banks do this? I've always wondered why we hear a lot about crypto exchanges getting hacked, but seldom about banks. What is it that banks are doing right (or crypto exchanges doing wrong) in terms of security?
mupuff1234 4y ago
Most people aren't bashing the tech, but bashing the excessive hype, fraudulent scams and money chasing.
- jiggawatts 4y ago
  It’s like… I can be fascinated by the creativity that goes into designing and making credit card skimmers that blend invisibly into an ATM, that doesn’t mean I like the theft… Ref: https://news.ycombinator.com/item?id=32843961 https://news.ycombinator.com/item?id=32843961
fladrif 4y ago
Having read your article, I still don't understand one part. The claim that the honest validators in the face of a malicious superminority can eventually leak them out, but a malicious supermajority cannot do the same to an honest superminority. I figure there would need to be some other mechanism that would tip the balance in favor of the honest validators, otherwise it seems like majority should always win.
- dannyw 4y ago
  If you run a node, you're also checking that every block follows the rules. A malicious supermajority cannot break the rules your client enforces, because your client will reject it. Think of it as everyone running ethereum has a bunch of asserts() each block or communication it receives.
  - fladrif 4y ago
    What are these rules? Say for instance 51% of validators decide to include a malicious transaction inside of a block, what rules would that be breaking?
    - capableweb 4y ago
      Depends on what way the transaction is "malicious". If malicious means transferring funds that don't exists, it'll be noticed. If malicious means transfer funds out of an address it doesn't hold the key for, it'll be noticed, and so on.
      - spywaregorilla 4y ago
        > If malicious means transferring funds that don't exists, it'll be noticed. who cares? As far as I can tell, a majority that wants to block a vote can do so freely and the only resolution is a fork where people just assume that the honest fork will win out. I also think it's not a majority but actually just a little more than a third to block a vote for eth
  - 4y ago
syzygyhack 4y ago
Ty for your service foo.
lucozade 4y ago
I have a question if I may. What mechanism is deciding who pays inactivity leaks? It says that if a minority stop attesting then they will leak until eventually the attestors get to a supermajority. That makes sense. It also says that, if a dishonest minority start a soft fork, the fact that they stop attesting on the honest fork means that they eventually leak out until the honest fork gets a supermajority. That implies that, unless some mechanism has decided which is the honest fork, then all attestors will leak assuming that they aren't trying to attest to both blocks (which is illegal). But if all attestors are leaking then that supermajority won't occur will it? So something must be deciding which is honest. But it can't be using number of attestors/deniers because of USAF ie where the honest fork is the less attested one. So how does that work? How is honesty determined given that both forks are legal wrt rules and failed attestation is penalised but cannot be shown to be malicious (according to the doc). Also, as an aside, how are leaks actually transacted? They can't be using the main transaction or none of the above would work. Is there some sort of shadow transaction system for staked ETH? If so, what mechanism decides the validity of the leak transactions? Very interesting mechanism. Clearly a lot of thought has been put into it.
- lucozade 4y ago
  In case anyone's interested, from what i can tell from reading other writeups, there is no "honest fork" check. Each fork will independently deal with inactivity until they're able to finalise. If that's the case, the article is correct in saying that a minority honest fork will finalise as would an honest supermajority fork. What it didn't mention is that the respective dishonest forks would too. If I've understood correctly. Not really sure I get why this is a good thing.

SilasX 4y ago

Guess you can update this part now! >3. PoS (Ethereum, soon™)

0xfoobar 4y ago
Indeed! :)

atomlib 4y ago

What's Substack? Is that new Medium?

speedylight 4y ago
That’s how I always thought about it yeah.
NavinF 4y ago
Yeah it's like Medium except it has less fluff/self-promotion.
lloppal 4y ago
It's like Medium but wants you to pay for newsletters.
- polisteps 4y ago
  Which is fairer imo than paying for the whole site. I prefer creator control on which content is paid and which content is not.
toastal 4y ago
It'll be okay like Medium used to be until it isn't and the funders turn it into Medium while seeking that return on investment. You're still better off owning your own blog in the long run.

bambax 4y ago

Isn't it a little unfortunate that the acronym for Proof of Stake would be PoS? Couldn't they think of something else?

kwhitefoot 4y ago
You mean people might confuse it with Point of Sale? /s
- capableweb 4y ago
  Don't be obtuse. They clearly imply people might think it means Point of Service.
- butterlesstoast 4y ago
  I don’t think that’s an obtuse comment. I understood the article to be talking about Point of Sale the entire time because I live off the blockchain hype.
  - Consultant32452 4y ago
    Hopefully the new algorithm is not a Piece of Shit.
  - kwhitefoot 4y ago
    I think you missed the joke. Or was it? Hmm, now I'm wondering!
kortex 4y ago
I'm sure you can find a naughty word or phrase, or some other semantic collision, out of almost any 3-letter combination, in some language somewhere. Sure it's one of the more well known ones in a very common lingua franca, but we humans are smart and can context switch. Edit: Actually not sure if you are talking about Point of Sale, Place of Service, Piece of $#!7, or something else. Case in point: context matters.

bsaul 4y ago

I find the complexity of that algorithm both impressive ( since they seem to have made it work), but also quite worrying. I'm really not sure how such a beast can't be filled with bugs, not in the implementation but rather in the protocol. I know a lot of very smart people are working on this, but i'd rather have something conceptually simpler to work as the base layer for a whole new economy.

thrown_22 4y ago
Hendrikto 4y ago
Making PoS scale to hundreds of thousands of nodes with commodity hardware is not simple though. Few projects managed so far, and Ethereum wasn‘t designed for it from the outset, so it‘s even more difficult.
- preseinger 4y ago
  But of course no system can scale to O(100k) nodes and retain reasonable availability and consistency properties.
- valzam 4y ago
  Of course they aren't really scaling to hundreds of thousands of nodes. At any given time only 120 nodes are involved in consensus.
arlcode 4y ago
Cryptocurrencies impose some complex constraints on themselves that require complex solutions. Conceptually banks and exchanges solved the consensus problem decades ago and they did it with a highly secured simple database and lots of crosschecks. But if you trust nobody (except some developers somehwere) then things get tricky
- chompychop 4y ago
  Could you go into more detail or provide references on where I could read up more on how banks do this? I've always wondered why we hear a lot about crypto exchanges getting hacked, but seldom about banks. What is it that banks are doing right (or crypto exchanges doing wrong) in terms of security?

mupuff1234 4y ago

Most people aren't bashing the tech, but bashing the excessive hype, fraudulent scams and money chasing.

jiggawatts 4y ago
It’s like… I can be fascinated by the creativity that goes into designing and making credit card skimmers that blend invisibly into an ATM, that doesn’t mean I like the theft… Ref: https://news.ycombinator.com/item?id=32843961 https://news.ycombinator.com/item?id=32843961

fladrif 4y ago

Having read your article, I still don't understand one part. The claim that the honest validators in the face of a malicious superminority can eventually leak them out, but a malicious supermajority cannot do the same to an honest superminority. I figure there would need to be some other mechanism that would tip the balance in favor of the honest validators, otherwise it seems like majority should always win.

dannyw 4y ago
If you run a node, you're also checking that every block follows the rules. A malicious supermajority cannot break the rules your client enforces, because your client will reject it. Think of it as everyone running ethereum has a bunch of asserts() each block or communication it receives.
- fladrif 4y ago
  What are these rules? Say for instance 51% of validators decide to include a malicious transaction inside of a block, what rules would that be breaking?
  - capableweb 4y ago
    Depends on what way the transaction is "malicious". If malicious means transferring funds that don't exists, it'll be noticed. If malicious means transfer funds out of an address it doesn't hold the key for, it'll be noticed, and so on.
    - spywaregorilla 4y ago
      > If malicious means transferring funds that don't exists, it'll be noticed. who cares? As far as I can tell, a majority that wants to block a vote can do so freely and the only resolution is a fork where people just assume that the honest fork will win out. I also think it's not a majority but actually just a little more than a third to block a vote for eth
- 4y ago

syzygyhack 4y ago

Ty for your service foo.

lucozade 4y ago

I have a question if I may. What mechanism is deciding who pays inactivity leaks? It says that if a minority stop attesting then they will leak until eventually the attestors get to a supermajority. That makes sense. It also says that, if a dishonest minority start a soft fork, the fact that they stop attesting on the honest fork means that they eventually leak out until the honest fork gets a supermajority. That implies that, unless some mechanism has decided which is the honest fork, then all attestors will leak assuming that they aren't trying to attest to both blocks (which is illegal). But if all attestors are leaking then that supermajority won't occur will it? So something must be deciding which is honest. But it can't be using number of attestors/deniers because of USAF ie where the honest fork is the less attested one. So how does that work? How is honesty determined given that both forks are legal wrt rules and failed attestation is penalised but cannot be shown to be malicious (according to the doc). Also, as an aside, how are leaks actually transacted? They can't be using the main transaction or none of the above would work. Is there some sort of shadow transaction system for staked ETH? If so, what mechanism decides the validity of the leak transactions? Very interesting mechanism. Clearly a lot of thought has been put into it.

lucozade 4y ago
In case anyone's interested, from what i can tell from reading other writeups, there is no "honest fork" check. Each fork will independently deal with inactivity until they're able to finalise. If that's the case, the article is correct in saying that a minority honest fork will finalise as would an honest supermajority fork. What it didn't mention is that the respective dishonest forks would too. If I've understood correctly. Not really sure I get why this is a good thing.