Decide whether to remove obsolete incremental test suite

[stefanmaierhofer]

Summary

src/Tests/Aardvark.Base.Incremental.Tests/Aardvark.Base.Incremental.Tests.fsproj is currently outside the normal validation path and appears to be effectively dead.

It is not included in:

• src/Aardvark.sln
• test.sh
• test.cmd

A retargeting attempt from netcoreapp3.0 to the current SDK/toolchain showed that this is not just a stale target-framework problem. The suite depends on a broad legacy adaptive/test helper surface that no longer matches the current repository.

What the retargeting attempt exposed

After switching the project to a current target framework and trying to reattach it to the standard test path, the failures spread across multiple areas:

• removed or renamed mutable helper types and collection aliases
  • CList, PList, CMap, MSet, MList, MMap, HSet, COrderedSet
• old adaptive set/list helper surface
  • ASet.toMod, ASet.toArray, ASet.mapM, ASet.flattenM, old callback helpers
• old reader/member assumptions
  • legacy reader method names and disposal expectations
• older Aardvark.Base API assumptions
  • e.g. V3d.Distance, AdaptiveObject.Id
• test support code that assumes older collection/runtime behavior

The breakage is distributed over files such as:

• src/Tests/Aardvark.Base.Incremental.Tests/AListTests.fs
• src/Tests/Aardvark.Base.Incremental.Tests/AListTestsNew.fs
• src/Tests/Aardvark.Base.Incremental.Tests/AMapTests.fs
• src/Tests/Aardvark.Base.Incremental.Tests/Mutables.fs
• src/Tests/Aardvark.Base.Incremental.Tests/Performance.fs
• src/Tests/Aardvark.Base.Incremental.Tests/SizeOfVariousAdaptiveObjects.fs
• src/Tests/Aardvark.Base.Incremental.Tests/UndoRedo.fs

Why this matters

Right now the repo gives the impression that this is still a real test suite, but normal CI and the normal local test entrypoints do not exercise it.

That leaves an ambiguous state:

• either these tests are deprecated and should be removed,
• or they still matter and need an explicit, broader migration plan.

Proposed decision

Please decide one of these explicitly:

1. Remove Aardvark.Base.Incremental.Tests entirely if it no longer provides value.
2. Keep it, but treat revival as a dedicated migration task instead of a small maintenance fix.

If we keep it, the follow-up should start by deciding which parts are still relevant and whether compatibility shims are acceptable, instead of trying to mechanically retarget the whole suite.