Clarifications on proof of work & phase out following Tuesday's governance meeting

#1

There’s been some news coverage following Tuesday’s governance meeting that have had some inaccuracies and confusing statements in them. It’s not always easy to keep track of what’s being said, what is meant, and what the actual implications are, even for those very close to the project. So I can understand that journalists struggle with this as well. As someone who attended the meeting, I wanted to make some comments to follow up on the discussion and hopefully make things a bit clearer.

  1. First of all, the canonical “source” of what actually is being discussed, said, and agreed in the weekly meetings (both development and governance) are the meeting notes. Whenever in doubt, go to the meeting notes and you shall (hopefully) find answers. Each note also references the actual Gitter discussion, where you can find verbatim and chronological quotes, unaltered. To make it easier to reference, a decision log is being kept up to date, listing specific decisions / agreements and where these can be found in the notes. If you ever encounter something that’s to your understanding is incorrect or misrepresenting in any way, please submit a pull request.

  2. Regarding proof of work, the relevant section in the last meeting is here.

  3. Now for the clarifications, which I’ll try to keep layman-friendly.

    1. Grin has committed to four scheduled hard forks, roughly six months apart from each other, to be used for network wide upgrades and to maintain ASIC Resistance for CuckARoo. The first one is targeted for ~ July 2019. This schedule remains intact and was not discussed in the meeting.

    2. The ASIC Resistant proof of work, CuckARoo, was not discussed in the meeting at all.

    3. CuckAToo is what we call the ASIC Targeted proof of work. Different “size variants” of this algorithm can mine concurrently on the network and are rewarded differently. Examples are CuckAToo 31 and CuckAToo 32, or AT31, AT32. A smaller number means a smaller size, which is a smaller difficulty, which gives you a smaller expected reward in proportion.

    4. The phase-out schedule, where smaller sizes gets phased out, is something that is programmed to happen gradually over time, in an automated fashion. A hard fork or a soft fork is not required for phase out to occur.

    5. In the meeting, we put together a tentative timeline of when each phase out would occur. It’s in the notes, but I share it here below for convenience. These are estimates.

      Cuckatoo Start of
      phase out
      Expected to be
      fully phased out
      AT31 Jan 15 2020 Aug 19 2020
      AT32 Jan 15 2021 Aug 27 2021
      AT33 Jan 15 2023 Sep 3 2023
      AT34 Jan 15 2027 Sep 10 2027
    6. In the meeting, there was not a suggestion or support for altering the phase out of AT31. It’s due to complete as planned by Aug 19 2020 or thereabout.

    7. In the meeting, there was a suggestion to commit to a fixed timeline of 18 months into the future, during which we would agree not to introduce any changes. Currently the commitment is to “the foreseeable future”, which is vague. If there’s would be such a commitment, it would mean that a theoretical change that gets agreed today, will take effect 18 months from now. Nothing was decided about this in the meeting, it will be discussed further in coming meetings.

    8. The phase out of AT32 is currently set to automatically begin on or around January 15 2021, ie in ~21.5 months from now.

    9. Over the next weeks and months, we may consider changing the timeline of this AT32 phase out. We don’t know. The only clear agreement in the meeting was to wait and see.

Happy to answer any further questions. If you see something that’s inaccurate above - please point this out. You’re all encouraged and welcomed to participate in the discussions.

1 Like
Obelisk GRN1 Full Sale