Firedrake meeting 2025 05 07 - firedrakeproject/firedrake GitHub Wiki

Date and time 2025-05-07 1600 UTC

Action Items

Pick Chair and Minuter (KS pick)
ALL: (ongoing) triage the open issues and confirm if they are indeed still open (and perhaps provide labels)
ALL: do things with SV's branches
DH: Email to Andreas to have 2 (+ others!!!) loopy PRs merged TODO: FIND OUT WHICH PRS THESE ARE
DH: Get Firedrake a docker open source account (link here)
DH: Talk to GregVernon about PR#2116.
JB: Enable merge queues (minutes)
LC: Try to merge RNH' PR: Movable VOM
PB: Profile and speed up some tests (minutes, minutes)
CW: More testing configurations (minutes)
All: Do post merge works (interpolate, abstract reduced functionals, etc.)
JHC, KS: Make docker and pypi accounts and ask CW to invite you.

Present: DH KS IM LC PB DH CW JHC

Apologies: RK

I would like to test a number of different configurations including:

What sort of testing coverage do we want for these? In particular, should we think about not running the entire test suite for complex mode?

We want to cover different builds, but running the entire test suite for everything is too big.
int64 we can check it runs but not that it works at >2billion DoFs
complex mode actually changes the maths being done so makes sense to run everything. Code changes also often break something in complex so we need to catch that early.
GPU needs special treatment, but only some functionality will be able to run on GPU anyway.
SuperLU and hypre will only run on GPU if built for GPU, so we need to check they are set up correctly in both cases.
We could make a github label for running the GPU CI like with macOS

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