Remove workaround that allows features to pass on Ruby 4.0

[faisal]

In 268d3a2 we introduced a workaround to avoid a Cucumber incompatibility with Ruby 3.5. When Cucumber addresses the issue we should remove (or more tightly gate) the workaround.

[faisal]

cucumber/cucumber-ruby-core#299 is tracking a possible Cucumber fix.

[faisal]

cucumber/cucumber-ruby-core#299 is tracking a possible Cucumber fix.

This does not fix the situation for our case. Or, at least, removing the workaround still results in failures.

[faisal]

I removed the monkeypatch as part of proving out #533 wouldn't block if ruby-head failed, but failed to get the monkeypatch back in before @etagwerker merged the PR.

Rather than re-re-rereverting, should we just live with ruby-head failing and fix it (along with other ruby-head issues) on the way to shipping Ruby 3.5 support? Or should I re-introduce the monkeypatch?

[etagwerker]

I removed the monkeypatch as part of proving out #533 wouldn't block if ruby-head failed, but failed to get the monkeypatch back in before @etagwerker merged the PR.

Oops, sorry about that.

Or should I re-introduce the monkeypatch?

Let's go with this.

[faisal]

I removed the monkeypatch as part of proving out #533 wouldn't block if ruby-head failed, but failed to get the monkeypatch back in before @etagwerker merged the PR.

Oops, sorry about that.

Or should I re-introduce the monkeypatch?

Let's go with this.

"this" as in what we have now, or "this" as in the suggestion to re-introduce the monkeypatch

FWIW the PR that cleared things for Features didn't fix the other test jobs, so I've done so in #534. If we want to reintroduce the monkeypatch we could do so there.

[faisal]

This may no longer be necessary. See cucumber/cucumber-ruby-core#299 (comment)