Testing of ANU-tub using different MOM versions

[chrisb13]

@jbisits has been able to run ANU-tub and I'm having some issues running mom6 standalone using a spack build / our infra. This issue (and related PR) is to test the different options, namely:

• angus-g / MOM6 -- this branch
• jbisits / MOM6 -- this branch
• https://github.com/ACCESS-NRI/MOM6/tree/2025.07.000. This already builds here.

Related:

• Running anu-tub with a different MOM6 executable AndyHoggANU/anu-tub#14

Useful resources for @jbisits:

• https://access-community-hub.github.io/access-iom3-configs/infrastructure/Building/#building-mom6-standalone-with-a-pre-release-build
• https://docs.access-hive.org.au/models/build_a_model/create_a_prerelease/

[chrisb13]

I’ve had a go at compiling using OM3 build CI. I actually started this as an example for why @jbisits and @angug-g might want to consider using our GitHub workflow build tools. It also means we can try out each other's executables, easily test reproducibility, see build logs etc.

From @jbisits

As far as I can tell merging 2025.07.000 into my branch worked but could be worth taking a look to see if that is the case? I can but not knowing the code base as well might be hard for to spot anything that might have been missed.
Let me know if there is anything else I can do! I am planning to make an OM3 build over the next few days with spack to test the online numerical mixing diagnostic. Will keep you posted on how this goes.

Okay, so I've tried building using Spack and both @jbisits and @angug-g's code; both fail. Specifically:

• here using 938539612332491a9ef9159990fc93c2a603efe5 or here
• here using 5a185920b1231849966ec35ee8f77ffe2f348d37 or here

Note that @jbisits is not taking the tip of the branch but a commit from 3 days ago (I missed the recent commit this morning it seems.)

[angus-g]

I expect my old branch won't build with CMake without at least telling it about filter_utils.F90. I guess you'll need to slot that into the pipeline somewhere?

[chrisb13]

We can close this now I think (see test example). Thanks @dougiesquire and others for all your work on this. fyi @jbisits.