Future structure/name of `ACCESS-Community-Hub access-iom3-configs`

[chrisb13]

@dougiesquire and @chrisb13 had a chat about this today.

We had a look at current candidates:

• https://github.com/ongqingyee/idealised-ASC/
• https://github.com/AndyHoggANU/anu-tub/

Considerations

• access-iom3-configs is a community space for people to use idealised versions of the OM3 component parts of models;
• subsets of the models in access-iom3-configs will be used for mom6 node testing and thus need to be able to use the ci repro infra';
• we discussed the idea of making forks but current examples (see above) arn't really close enough to being payu runnable for it to make sense;
• pre-building mom6standalone would help people like @jbisits and @edoddridge who in some cases don't want to change code (see below solution);
• would prefer not to use submodules.

Based on the above, things to do:

• add big disclaimer that the contents of access-iom3-configs in some cases only includes one of the components;
• discuss next Monday and at TWG about putting a MOM6 standalone executable alongside our other executables;
• consider using folders inside access-iom3-configs not branches (@dougiesquire's strong preference -- like what is done here);
• talk to @AndyHoggANU (DONE)/@angus-g and
• eventually @ongqingyee/@navidcy about putting a copy of one of their configurations into this repo;
• eventually https://github.com/ACCESS-NRI/model-config-tests may need to be modified to work. Let the release team know that this is the plan.
• ask @jbisits about whether he got anu-tub working with a spack built mom6 access-nri fork executable not just Angus' one;
  • depending on the above, @dougiesquire to have a go next week at using access-om3 built executable to run anu-tub.

Feel free to edit this post if I missed anything @dougiesquire.

fyi @jbisits @helenmacdonald

[chrisb13]

@jbisits. The above chat was productive and good news for access-iom3-configs and your work (both in the #10 sense and your numerical mixing tests). We'll give the highlights on Wednesday at the TWG if you'd like an update (alternatively, if you're time poor, I can just steer us in the appropriate direction).

[jbisits]

@jbisits. The above chat was productive and good news for access-iom3-configs and your work (both in the #10 sense and your numerical mixing tests). We'll give the highlights on Wednesday at the TWG if you'd like an update (alternatively, if you're time poor, I can just steer us in the appropriate direction).

Great! I will come along on to TWG on Wednesday. Is it at 3pm?

[jbisits]

I have a spack build of mom6 from 2025.07, and a build using mom6-ninja-nci, and I was able to use both to run ANU-TUB. Let me know if you would like more details. If you are a member of project e14 I can post the path and you can look at it there.

[chrisb13]

Great! I will come along on to TWG on Wednesday. Is it at 3pm?

Yep. Here's a draft agenda/details, feel free to add to it.

I've added you to next year's series but they're still running this year -- it's just @dougiesquire's old invite for this year.

[chrisb13]

[ ]  talk to @AndyHoggANU/[@angus-g](https://github.com/angus-g) 

Spoke to @AndyHoggANU about this over lunch and he's okay for us to put whatever files are needed into this repo' this includes the script that was given to @jbisits.

@angus-g are you also okay for us to put the anu-tub files into this repo for testing and idealised model config purposes?

fyi @dougiesquire

[chrisb13]

On this

Let the release team know that this is the plan

Meeting organised for tomorrow.

[angus-g]

@angus-g are you also okay for us to put the anu-tub files into this repo for testing and idealised model config purposes?

fine with me!

[chrisb13]

Meeting summary with @dougiesquire @anton-seaice @CodeGat @helenmacdonald @chrisb13 from this afternoon.

@aidanheerdegen was an apology.

Background...

As part of mom6 node testing and chris' passion project (helping folks use/run idealised test cases) there's been some progress on how to test and build mom6 in idealised setups. Eventually we'd like this to include testing of other om3 components

There's some background here:
 - #11 (comment)
 - #ocean-seaice > Pipeline for MOM6 node testing

And a PR proof of concept here:
 - ACCESS-NRI/ACCESS-OM3#166

Anton wanted us to check in with the release team (i.e. Is it okay to have MOM6 standalone in the access-om3 MDR). He's concerned about maintenance burden and scope — have I got that right @anton Steketee?

Decisions for the meeting include:
 - do we think it's okay to have a mom6 standalone build in the existing access-om3 mdr?

Outcomes/next actions

@_Harshula|784233 said:

TODO:
* Dougie: Proof of concept of (a) the ACCESS-OM3 SBR depending on MOM6 Standalone and (b) modification of MOM6 CMakeLists.txt to build both the library and the executable together.
* Chris: Present the Proof of concept and persuade Aidan that it is a good idea.

@_Harshula|784233 said:

During the ACCESS-OM3 w/ MOM6 standalone discussion with @**Chris Bull** , @**Dougie Squire** , @**Anton** , @**Tommy Gatti** , the following nice to have `build-cd` features were mentioned:

Could `build-cd` display the following in a comment:
* Foldable Spack concretizer output
* Foldable spack.yaml content
* Relevant dependency versions containing hyperlinks to source code

Tommy will come along to consider possible options above at tomorrow's TWG (so won't do anything until it's been discussed). (Chris has mentioned this to @aekiss who has also been interested in ideas like this.) Also relevant to the upcoming COSIMA training.

Further chat here:

• #model-release > build-cd @ 💬
• #ocean-seaice > Pipeline for MOM6 node testing