Page tree
Skip to end of metadata
Go to start of metadata


OutcomeKeep TSC term length to 1 year.
Minutes Link 2019 10 10 TSC Minutes

Should the TSC move to terms of n years, with 1/n of the members being elected each year?


  1. Proposal: Extend to 2 years. To get started, randomly select half of the members (rounded down) whose term gets bumped to 2 years immediately.

  2. 2 years with 50% up for election each year sounds like a good plan.

  3. I'm not so sure about this.  We've had TSC members basically fade out because their job responsibilities changed (or they totally changed jobs).  No one seems to want to resign, so we essentially get non-participants.

    I guess the argument for 2 year terms would be more continuity on the TSC.  But it seems like we already have that, so I'm not sure what problem this is solving.

    Arnaud or Gary, can you persuade me why this would help us?

    1. Perhaps extend the other proposal about losing your vote? perhaps we have a way for TSC members to "vote someone off the island"?

    2. I think having some type of process whereby members can be removed for non-participation would be the way to solve this if we really get into a situation where we are not getting the aggregate level of participation we want.  In Fabric we have a stated policy on voting to retire maintainers and it seems to work pretty well.

      The main reason I like the 2 year term is because I think it can take new members a while to get up to speed plus I also think it's good to have continuity across the membership (assuming that someday we don't have as many repeat members).

  4. I was thinking about this too. I think 50% each year would be good for continuity and also in my experience it took me a good chunk of the year to really get oriented and tuned in to the longer term debates/discussions.

  5. I like the idea. Coupled with the quorum attendance requirement and strengthening that with a non-participation disqualification clause (non-excused) of say 2 months might just do the trick.

  6. Any time we add process, we add rules that have to be remembered and followed, and that creates opportunities for errors and then subjective decisions made to correct for those errors.  Governance models should be as simple as possible, but no simpler, to paraphrase Einstein.  We saw 45% turnover this year, and no sign that such turnover means that the TSC will be any less effective this year than if it had seen 0%.  Let's not create new processes or solutions for problems it's not well established that we have.

    1. I am convinced by this argument.

      Of course if we make the rules sufficiently interesting to automate we could run the governance process on-chain...

      1. thus fulfilling the prophecy "one chain to rule them all"....

        1. and in the darkness, bind them