Proposal 1: Working groups should be scoped to a specific task or problem rather than a general area.

Currently working groups are focused on a fairly general area. We propose that the scope of working groups be limited to a specific problem or technology.

7 Comments

  1. Christopher Ferris

    I'm starting to warm up to Hart's characterization as a Task Force rather than a Working Group. I would also like to see that we scope the lifetime.

  2. Hart Montgomery

    Mic gave a great example of what I think a task force should be this morning.  Imagine a group of people across projects wanted to investigate shared consensus.  They could create a task force whose stated goal was to investigate the possibility of creating a lab for shared consensus.  The project would be time-limited and the ideal goal would be to actually come out with starter code for a consensus lab.  If it worked, great, if it didn't, that's OK too, but it would be over relatively quickly either way and things wouldn't drag on forever.

  3. Christopher Ferris

    Agreed, but IMO, the creation of such should not be something requiring TSC approval. I do think that it should warrant notice to the TSC mailing list, and have a Wiki page as this does. I also think that it should report its progress and conclusions, but that shouldn't add significant process overhead as to be a disincentive.

    1. Mic Bowman

      I definitely see your point... And frankly, I think we're really just proposing that we formalize "task force". Though I think, to echo Shawn Amundson it would be good to have formal documentation & tracking. Think "report out" from a completed task force.

      And... Just to provoke discussion... It seems to me that there is a big difference between an HL Labs project where the lab owner is really doing a single, one-time dump of code and a Labs project that is intended to be multi-organization pre-work for either a full fledged project or some new component of an existing project. FPC, for example, is being run with a very specific eye towards creating something that enhances Fabric (could be rolled into Fabric, could be a sub-project, could be a separate maintained component). I think URSA is another great example. 

      There is a clear difference between resource requirements for something like Ursa and FPC (where we actively use or want to use all the facilities that are normally given only to a full-fledged project). TSC approval could be given for limited time access to those resources.

      Finally... i think WGs like this could be the "technical" tool for the TSC. Again... thinking of a more formalized task force, a TSC that was acting "intentionally" might identify a gap (e.g. a common standard for externalizing ledger state) and create a task force to address that gap. 

      Bottom line for me... WGs go away. The discussion part moves into more lightweight SIGs. The operational part becomes a better documented, single purpose/well-focused task force.

      1. Hart Montgomery

        +1 to this.

        As an aside, we should probably "garden" the labs more.  If they don't get any updates over a certain amount of time, then we should take some action to indicate this.

        Mic correctly points out some of the issues we had with Ursa as a lab.  We had a lot of communication and wanted proper communication and meeting channels.  We were (probably illegally, at least according to the rules at the time) given a chat channel and a meeting slot.  We used the labs email list, and this only worked because we were really the only people that used it.  Giving other "active" labs the tools to do this would be a very useful thing, in my opinion. 

        "Bottom line for me... WGs go away. The discussion part moves into more lightweight SIGs. The operational part becomes a better documented, single purpose/well-focused task force."

        I agree 100%.  

        I'll point this out though:  well-run discussion groups would be likely spawn some of these new task forces.

        1. Mic Bowman

          To this point. Should we just have one proposal which is that working groups go away? I think, if we do that, we probably need to create a more substantial description of a task force. 

          Or we just turn working groups into a focused task force?


  4. Christopher Ferris

    I think that the gardening of the Labs should be the function of the Stewards, no? To Mic's point, I tend to agree that there is most definitely a spectrum of Labs projects ranging from single author tool/utility/sample to collaborative experimentation that might start to grow legs and arms. Projects like FPC probably should graduate to incubation when there is clear sense that there is a growing interest and community of contributors that span more than one company.