Souptacular commented 6 years ago

Ethereum Core Devs Meeting Constantinople Session 2 Agenda

Meeting Date/Time: Friday 19 October 2018 at 14:00 UTC

Meeting Duration 1 hour

Agenda

Ropsten issues.
When fork?

5chdn commented 6 years ago

Maybe a good idea to collect all the issues here, so everyone is on the same page and we can focus on resolving problems rather than talking about what happened on the call:

Meta issues

recently added hashpower caused reduced blocktimes and caused this hardfork to happen much earlier than expected on a Saturday which is by all means the worst time for a hardfork
hardfork happened only 6 days after Geth release and 1 day after Parity Ethereum release, users had not enough time to upgrade
not a single miner was mining on the Constantinople chain, it took us two hours to find someone to sync an updated Ropsten node and start mining 4_230_000 while Byzantium chain was happly grinded at ~1 GH/s
there is no fork monitor available, only http://ropsten-stats.parity.io which does not reveal details about the different chains

Client issues

Harmony used two different configs for Ropsten, see https://github.com/ethereum/ethereumj/pull/1214
Parity Ethereum consensus bug in EIP-1283 implementation, see https://github.com/paritytech/parity-ethereum/pull/9746
Aleth similar bug in EIP-1283, see https://github.com/ethereum/aleth/pull/5316

Anything else?

Edit: Lane moved this to the forum, makes sense: https://ethereum-magicians.org/t/issues-we-discovered-in-the-ropsten-constantinople-hard-fork/1598

lows93 commented 6 years ago

will we go back to the future ?

`Meeting Date/Time: Friday 31 August 2018 at 14:00 UTC´

lrettig commented 6 years ago

Fixed the date

Souptacular commented 6 years ago

lol @lrettig we fixed it at literally the exact same minute and GitHub errored out for me.

lrettig commented 6 years ago

Good to know consensus by GitHub works so well 👍

lrettig commented 6 years ago

This is my attempt to summarize the things that went wrong - and by extension, things I'd love to discuss on Friday's call. Please make additions/corrections! https://ethereum-magicians.org/t/issues-we-discovered-in-the-ropsten-constantinople-hard-fork/1598.

@5chdn - thanks for getting the list started above. Mine includes most of the ones on your list, with slightly more focus on "process" and less on code. We can try to merge them into a canonical list if you like. I have a slight preference for keeping threaded conversation on the FEM forum and using Github more for high-level process, call scheduling, etc.

5chdn commented 6 years ago

I approve this thread, Lane, and I appreciate that you collected all these items. There are two differences though:

I tried to order my items in a chronological manner.
I tried to focus on all the things that are immediately important for the call.

But it totally makes sense to also go through the additional questions you raised, but maybe not all in one call :rofl:

AlexeyAkhunov commented 6 years ago

I will not be able to join the call. I have read @lrettig thread on Magicians. What is missing there and also here: we need to answer the question: "What is the plan between now and mainnet hard-fork?". Clearly, we are not rolling back Ropsten, and keep testing things there. So I propose regularly spawning (and mining) temporary test nets to test transition to Constantinople with any clients that want to participate. It is important to test both the new code (on Ropsten now), and the tools and processes used for transition (not possible on Ropsten anymore).

AlexeyAkhunov commented 6 years ago

Oh, and also forgot another suggestion - what about bounties for causing consensus issues on Ropsten?

egalano commented 6 years ago

I'd like to propose bounties for running nodes or mining on Ropsten as well to encourage more community participation. At the very least during the periods leading up to a fork.

5chdn commented 6 years ago

we need to answer the question: "What is the plan between now and mainnet hard-fork?"

Tests! :)

eggy-eth commented 6 years ago

How about someone also writes down the main points which are discussed in the call? I would really like to know, for example, what the consensus is on the second point. I know I can also listen to the call again, but this means I have to spend a hour to listening to the video to farm out all the points which have to be raised. I really think it would help more people as well to quickly get a "tldr" of the call.

EDIT: Eggy should read. It's already there, except not this meeting.

lows93 commented 6 years ago

So tldw;

Mainnet Hardfork not before January / End of January
time between release Hardfork-Client and the Hardfork itself should be atleast 4 Weeks
possible Progw for the upcoming Hardfork
for Istanbul and further Hardforks there should be a Testphase on Testnet for min. half a year

lrettig commented 6 years ago

Notes are up: https://github.com/ethereum/pm/blob/master/All%20Core%20Devs%20Meetings/Constantinople%20Meeting%202.md

olalawal commented 6 years ago

If progpow is a definate go for implementation in the delayed fork.. Me and my conglomerate of freelance miners are happy to set up a pool on Testnet and provide testing. We currently run a BCI progpow pools and Yiimp as well.

our discord: https://discord.gg/7VRe2aW

If ProgPow is not going to be deployed as part of the implementation we will not be interested in testing .. most of our miners are mining xmr , raven, bci .. asic/fpga proof algos.

xmr just forked a few days ago , and shed around 20 % of fpga hash

ethereum / pm