feat: MambaInstall plugin

[wietzesuijker]

Enables the use of both Mamba and Micromamba for package installation, offering flexibility depending on user preference or environment constraints.
Mamba accelerates package installation and dependency resolution compared to Conda.

[GPUtester]

Can one of the admins verify this patch?

Admins can comment ok to test to allow this one PR to run or add to allowlist to allow all future PRs from the same author to run.

[wietzesuijker]

ok to test from my perspective. Nit: sorry I initially missed the section on code formatting

[github-actions]

Unit Test Results

See test report for an extended history of previous test failures. This is useful for diagnosing flaky tests.

    27 files  ± 0      27 suites  ±0   10h 0m 35s ⏱️ - 9m 3s
 3 994 tests ± 0   3 884 ✅ + 1    109 💤 ±0  1 ❌  - 1 
50 230 runs  +18  47 939 ✅ +26  2 290 💤  - 6  1 ❌  - 2 

For more details on these failures, see this check.

Results for commit 40e5c88. ± Comparison against base commit b03efee.

[wietzesuijker]

Just realised that the channels arg could be squashed into the options. However, I'd rather have it explicit (if we go ahead with that, this would also need a small change in the CondaInstall class)

[jacobtomlinson]

Thank you for taking the time to raise this.

Given that libmamba is now the default conda solver it sort of feels like mamba will gradually disappear. Especially given the mamba devs are now focused on pixi.

What benefit does this plugin give over the existing CondaInstall with modern versions of conda?

[wietzesuijker]

Thanks I didn't fully appreciate those changes. I quickly tested CondaInstall and MambaInstall, and they perform quite similarly indeed! The only use case would be for setups that do have (micro)mamba, but not conda installed. Fine to close though!