Hardfork Meta EIP-1679: Istanbul discussion

Do you know who is going to be working on that change? If there is no one who wants to do that, the change will definitely not happen, and in that case there is no point of putting the EIP in.

There is no one. A bit stupid of me to assume I could just propose an EIP when no one is available to work on it.

As a sidenote - the EIP states that “today” (Friday 17 May) is the “hard deadline” to propose EIPs for Istanbul. A timezone is not mentioned, so it is not clear when this hard deadline actually is. I think it should be more clear what the actual time is regarding timezones?

We don’t need to be that strict about it to specify time zones.

No, it is not stupid at all. It has been the modus operandi before, but now I am trying to get the process participants accept the obvious (that if noone wants to do an EIP, it won’t be done), and also to remove the bottleneck of relying on go-ethereum or parity or Aleth developers to implement EIPs, as described here.

2 Likes

I am in the processing of being hired for testing coordination and helping dimitry. Would be great to sync up on good ways to approach this so I can start opening lines of communitcations with everyone needed. At least we will have one person able to focus on this as 25 EIPs is a lot to sift through. Also, I imagine testing requirements will be a good sifting point.

2 Likes

You should connect with the testing working group

2 Likes

Yeah I am talking with them.

EIP-1702 and EIP-1803 has no discussions-to

Here’s a spreadheet summarizing the scope of the proposed EIPs. - https://docs.google.com/spreadsheets/d/1evVKtEwEQwy4Gr84RjOpuf4CnjoISuB3865mvrXl-oc/edit?usp=sharing

Thanks, this look great!

I think the proper list of pending proposals is the following though: https://github.com/ethereum/EIPs/pulls?q=is%3Apr+is%3Aopen+label%3Aistanbul

Your spreadsheet seems to miss some of them.

I went through both and made a Readiness Checklist.

Steps for Checkpoint 1

  • Submit PR
  • Formatted Correctly
  • Received Review
  • Merged as Draft

Chart Link

I will go through and update more. Would love feedback on milestones to add to the checklist?

What does this stage means?

Basically means it builds and passes the formatting checks the bots do. Also, resolved any formatting requests on the PR.

After I go through and update the rest of the EIPs I think it will tease this out a little more. I know you have been putting a ton of work into this @axic :bowing_man:

I managed to pull the mergeable_status from the GitHub API Directly.

So at least checking if it builds and is mergable can be automated

1 Like

This is fantastic James! Nice work!

I’ve added this link to the wiki page https://en.ethereum.wiki/roadmap/istanbul

Re: milestones – I think if we can use this as the progress tracker that would be interesting. Less about the EIP spec and more about “implemented in” and then links to PRs / issues / commits in the clients.

Since the “hard deadline” for proposals came and went, what’s next for us champions? How do we know what is required of us moving forwards?

The short version: code reference implementations and schedule yourself into an ACD to get to Accepted.

Tests would be nice.

I think that’s about it? I mean, the other part I can think of is connecting with client teams on if they want to code their own implementation or if they want you to submit a PR.

What am I missing?

1 Like

Perhaps there is a special case for mine, since 1344 and 1965 might conflict.

More than willing to work on implementations in geth and parity for it.

For me the biggest indicator that an EIP should be accepted would be for someone to “Champion” it on this weeks or the next, 7 Jun, AllCoreDevs call. There are so many EIPs (especially EIPs submitted at the buzzer) that if a champion doesn’t step up on the call and gets immediate consensus I think the EIP should be bumped to “Asiago.”

1 Like

If that is the expectation we need to reach out and contact all the champions. I’ve sort of been thinking about this and @tvanepps helped put names to a number of them on the wiki.

Those that are officially under a working group should ideally be grouped and labelled as well.

Let’s use @MadeofTin’s spreadsheet to add some extra data like this.

1 Like