Ethereum Not Forking Over Asics, Casper Coming in Around Five Months

0

Sounds like the consensus of core developers is that we should not do anything at this time regarding the asics fork proposal, Hudson Jameson, an ethereum developer, stated after a public discussion between ethereum developers.

This is not bitcoin, miners are not in control, if they get evil, we’ll just speed up Casper’s development, Vitalik Buterin said.

The Bitmain asics miners is only 2.5 times more efficient, Buterin argued. It looks like it’s just a more efficient computer of sorts, he said. That means if eth forks the algorithm would have to be something which is not I/O bound, like sha3.

However he expressed concerns over the time it would take to implement that algorithm and whether it would delay other developments. Strongly lean towards no action, he said.

Casper FFG, or hybrid Proof of Work (PoW) / Proof of Stake (PoS) Casper is at the second stage of testnet, Buterin said. Totally ready for Geth/Parity to start implementing.

It is in the process of being verified and checked, but if 51% concerns demanded we can just skip audit and verification stages, Buterin said.

If Geth and Parity make it a high priority to implement Casper FFG, that gives insurance, Buterin argued.

The verification process would take around 5 and a half months, Danny, an eth dev, said. It started around mid-march, he said. There has been solid progress, daily.

Nick Johnson, another eth dev, said he does not think the current situation justifies a fork on its own, but it could be roll out into the next planned fork.

Definitely agree with that, Buterin said, arguing a fork on its own just for asics was not justifiable but could potentially be included in a planned fork.

Sounds like the consensus of core devs is that we should not do anything at this time, Hudson Jameson concluded after not one dev spoke in favor of a proposed asics fork.

The developers further discussed coding an asics resistant algorithm, with Buterin stating maybe as backup but in parallel with other work so that it doesn’t slow down other work.

In regards to whether asics would have an effect on hybrid Casper, Buterin said a 51% attack could prevent the finalization of new blocks, as in check-points, but can not break finality, as in blocks already check-pointed.

We can add a fork-choice rule to Casper where it should be prefered if it has a larger number of Casper/PoS votes, Buterin said.

So that if the attacking mining chain refuses to include, but good guy chain includes, that should make a 51% attack very week, he said.

The developers made it clear that if nonetheless the wider community insists on such fork they could make a proposal, but the focus for now seems to very much be on getting Casper FFG out, which might go live as soon as this Autumn.

 

Leave a Reply

Be the First to Comment!

100000
  Subscribe  
Notify of