Firedrake meeting 2025 03 05 - firedrakeproject/firedrake GitHub Wiki

Date and time 2025-03-05 1600 UTC

Action Items

Pick Chair and Minuter (DD to 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)

Agenda

Present: DH, AO, JHC, DID, IM, LC, KS

Apologies: CW, PB

CW: Switch to being root user in the Docker containers

From Francesco Ballarin:

And on the topic of docker: can anyone remind me what is the rationale for having a firedrake user and a local installation in $HOME/.local, rather than a simpler setup that uses directly the root user and installing everything in /usr? I do understand that outside of a docker image we'd never use the root user for security reasons, but the docker container itself is isolated from the rest and, if security was a concern, the current firedrake user isn't really secure either (since it must allow to sudo without password). I also do understand that in the previous iteration of docker images there was a firedrake virtualenv which probably justified a standalone user. But now, with that virtualenv gone, is there really a point in installing to $HOME/.local rather than the system wide path?

I agree with him. It causes needless complication (like having to set PYTHONPATH on GitHub-hosted runners) to get things to work properly. This is causing issues for some users (example).

What do people think?

DH + JHC: Can probably try and see if anything breaks.