Add support for numpy 2.0

[johnwparent]

Changes:

• [workspace] Upgrade pybind11 to latest commit (Replace removed itemset method pybind11#72).
• Disable Gynmasium-related regression tests on macOS. DrakeGym might not work anymore on macOS.
• Update some unit test syntax for latest numpy.
• Revert our temporary pin to numpy < 2. We no longer constraint numpy at all.
• Fix namedview repr to add <> to comply with Python's required semantics.

Towards #21577.

---

This change is 

[johnwparent]

cc @BetsyMcPhail

[jwnimmer-tri]

Thanks, looks promising so far!

The next step is to push a second commit onto this branch that reverts #21582, and then launch every macOS Unprovisioned Experimental build (to see how it works versus numpy 2).

[johnwparent]

@jwnimmer-tri pushed the revert, how do we kick off the extra jobs?

[jwnimmer-tri]

Docs at https://drake.mit.edu/jenkins.html, but here it is:

@drake-jenkins-bot mac-arm-sonoma-unprovisioned-clang-bazel-experimental-release please
@drake-jenkins-bot mac-arm-ventura-unprovisioned-clang-bazel-experimental-release please
@drake-jenkins-bot mac-arm-ventura-unprovisioned-clang-cmake-experimental-packaging please
@drake-jenkins-bot mac-arm-ventura-unprovisioned-clang-wheel-experimental-release please

[johnwparent]

Thanks!

[jwnimmer-tri]

@drake-jenkins-bot mac-arm-sonoma-unprovisioned-clang-bazel-experimental-release please
@drake-jenkins-bot mac-arm-ventura-unprovisioned-clang-bazel-experimental-release please

I pushed some unit test fixes. We'll see how new builds fare.

If the Gym tests are still broken with upstream bugs (they were yesterday), then we'll just disavow Gym support on macOS and disable them.

[jwnimmer-tri]

... then we'll just disavow Gym support on macOS and disable them.

I've pushed this change now too.

Remaining work is to fix the namedvector test failure, and get our macOS CI machines up an running again.

[jwnimmer-tri]

I'm going to let the current CI job finish to see where we stand, and then push the pybind11 git sha bump.

If the only remaining CI failure is the namedview test failure, I'll also push a nerf for that test.

Conceivably this will be green and ready to merge tomorrow morning.

[jwnimmer-tri]

Reviewed 3 of 3 files at r2, 2 of 4 files at r3, 2 of 3 files at r4, 3 of 3 files at r5, 3 of 3 files at r6, all commit messages.
Reviewable status: LGTM missing from assignee jwnimmer-tri(platform), needs at least two assigned reviewers, labeled "do not merge", commits need curation (https://drake.mit.edu/reviewable.html#curated-commits) (waiting on @johnwparent)

[jwnimmer-tri]

@drake-jenkins-bot mac-arm-sonoma-unprovisioned-clang-bazel-experimental-release please
@drake-jenkins-bot mac-arm-ventura-unprovisioned-clang-bazel-experimental-release please

[jwnimmer-tri]

I am somewhat hopeful that CI might pass this time around. In any cases, we're close.

+@EricCousineau-TRI are you able to feature-review this? Goal would be to land it before Monday.

[EricCousineau-TRI]

feature
+@ggould-tri for platform please!

Reviewed 3 of 3 files at r2, 2 of 4 files at r3, 2 of 3 files at r4, 3 of 3 files at r5, 3 of 3 files at r6, all commit messages.
Reviewable status: LGTM missing from assignee ggould-tri(platform), labeled "do not merge", commits need curation (https://drake.mit.edu/reviewable.html#curated-commits) (waiting on @johnwparent)

[ggould-tri]

Reviewed 3 of 3 files at r2, 2 of 4 files at r3, 2 of 3 files at r4, 3 of 3 files at r5, 3 of 3 files at r6, all commit messages.
Reviewable status: labeled "do not merge", commits need curation (https://drake.mit.edu/reviewable.html#curated-commits) (waiting on @johnwparent)