Firedrake meeting 2025 02 12

Date and time 2025-02-12 1600 UTC

Action Items

1. Pick Chair and Minuter (DD to pick)
2. ALL: (ongoing) triage the open issues and confirm if they are indeed still open (and perhaps provide labels)
3. ALL: do things with SV's branches
4. DH: Email to Andreas to have 2 (+ others!!!) loopy PRs merged TODO: FIND OUT WHICH PRS THESE ARE
5. DH: Get Firedrake a docker open source account (link here)
6. DH: Talk to GregVernon about PR#2116.
7. JB: Enable merge queues (minutes)
8. LC: Try to merge RNH' PR: Movable VOM
9. PB: Profile and speed up some tests (minutes, minutes)
10. CW: Fix artefactsv3 issue (done by DD for pyadjoint, no other instances found)
11. CW: More testing configurations (minutes)

Agenda

Present: PB, DH, KS, DD, CC, JHC, RK, CW, IM, LC

Apologies:

RK: Firedrake USA

RK: $1200 minumum for the buffet, get an invoice from Imperial if registration cannot cover that

RK: Friday morning tutorial: Q & A for newcomers

DH: We also want to devote some time to experienced users

RK: Saturday night plans, if we are 15-20 Rob can host us at his home, there is Uber for transportation. Some people will rent cars.

DH: we can still organize an unofficial dinosaur excursion if the weather is dry enough.

RK: we still need to schedule the talks. Registratioin deadline is Feb 13 (tomorrow). Abstract submission is Feb 17.

DH: website has been down for the last two weeks. CW: change actions to @v3.

CW: Finally remove .split() merged

#4204

The main controversy is renaming .subfunctions to .subspaces for function spaces. I think this was just a silly mistake we made when deprecating .split().

We don't want to merge it now as we want to announce it on Slack first and then merge later this week.

JHC: EnsembleFunction

New set of classes for representing finite element spaces/functions distributed over an Ensemble. Provides some collective semantics and operations that are either useful (e.g. dealing with one EnsembleFunction rather than lists of Function on each ensemble member), or downright necessary (e.g. _ad_dot to be able to create ReducedFunctionals over an Ensemble that actually passes the Taylor tests).

Still in draft form, I'm just looking for any feedback at this stage. See the description in the PR for more detail.

DH/CW PETSc updates

#3997 before #3999. Note that we expect the other workflows to fail because they are using the wrong PETSc (from the container).

DMSwarm API changed, we need to adopt the new API DMSwarm_cellid is now a variable that has to be accessed

We put our options and then pull them out, this incorrectly raised "unused options" warnings. Options which have not been used from our solvers are only those for which we raise warnings. Failing tests with large number of solver options. Needs more work.

CW: Gusto smoke tests

#4017

We should let other downstream packages know that we have done this so they can also submit a small set of tests to run.

CW: Time limit is 5 min. Takes 2 min.

CW: We want to do this for thetis, Irksome, ...

Merge PRs

Note that PRs put in this section should either be trivial or already have been reviewed. Discussion-worthy PRs should be separate agenda items.

DD: #4023 Move firedrake_adjoint tests from pyadjoint to firedrake/tests/firedrake/adjoint. merged

DD: #192 Remove firedrake_adjoint tests from pyadjoint. merged

PB: FEniCS-style bcs merged

PB: Subclass LinearSolver Reviewed half of it, DH concluded that this is adjoint-safe

Date of next meeting

1600 UTC 2025-02-19