Haskell Foundation Technical Task Force Meeting
26 February 2021
Present: Andrew Lelechenko, Théophile Choutri, Emily Pillmore(Chair),Chris Dornan, Ben Gamari, Andrew Boardman, Ed Kmett, Davean Scies
Proposal from Chris:
- Split up into parallel tracks focusing on particular areas of focus (GHC vs. Docs vs. etc)
- Not hermetically sealed teams, but rather a focus group to generate proposals for later meetings
- Consider what role HF should play: support? docs/ resourcing? Development? These should be part of the proposal.
Ed raises a point: Centralized design does need to be considered. We should make sure to avoid the hodgepodge that could result from random proposals.
Response: HF will be the one to locate leads and manage execution of these proposals as a project lead or resourcer for leadership on proposals.
Ben: GHC Steering committee has a proposals system to help with this as well.
Emily: A proposal process could help us in this, but that gets closer to a technical advisory board. Let’s punt. Andrew and I will discuss.
Ed: How do we shave off the rough edges to reveal a story about the current state of and future state of use for GHC?
Davean: Isn’t the mutable state constructors proposal stalled? (sebastiangraf is understaffed - no bandwidth)
Focusing on process:
Asynchronously, if we are happy that we’ve exhausted our wishlist, we’ll partition off to smaller teams to discuss concrete proposals, submit them to the group for nitpicking, and eventually to the board for ratification.
Davean: let’s not aim for too much structure.
We should feel free to bring in SME’s for particular tracks to influence thought on proposals.
First delivery of items for proposal by end of april.
Focus on underpromising + overdelivering. We want to focus on 3month to 1-year projects that we can get wins for in the short term to build trust and confidence that we can take on larger projects.
We probably do not want to initially commit to long term projects just to make sure we’re not taking resources away from the wins.
Chris: we need a looser coupled project structure to make sure we’re not committed to a long grand agenda. (cut down on bikeshedding and deliver instead of navel-gazing)
Ben: Decide on a set of things as a tight group, in the future, open it up and come up with a proposal process which is looser.
Davean: we also need to consider budget proposals
Davean: low-resourced projects for long term agenda items would be fine after perhaps a year
Ed: Num is a difficult problem that is not the best use of HF’s time.
UTF-8 Text is possible! We’ll consider it as an agenda item going forward.