deprecation updates ahead of 9.0.0

[davidjgoss]

🤔 What's changed?

• Some rewording and fixing around how we describe deprecations
• Undeprecate defineStep - at the time the direction seemed to be to eventually make strict keywords the only behaviour. However there is not a consensus across platforms for doing that, so on that basis it should only be an option. The PickleStep model has an 'Unknown' value so we can use that for defineStep and just throw if you try to use it with strict keywords. Finally, a user raised a separate concern in defense of defineStep from a linguistics perspective. So overall I think there's no longer a case to remove this.
• Deprecate the export of JsonFormatter class - currently most of the built-in formatter classes are exported like this but I think they should mostly be considered implementation details. When we tackle https://github.com/cucumber/cucumber-js/issues/2239 we'll rewrite this formatter, probably in the new "plugin"-style format we've discussed, so we should get this off the entry point as soon as we can.
  • Whilst it would be good to move early on this, we can't reasonably deprecate something without pointing to the new thing that people should migrate to, which we don't have yet, so omitting this for now.

🏷️ What kind of change is this?

• :bank: Refactoring/debt/DX (improvement to code design, tooling, documentation etc. without changing behaviour)

📋 Checklist:

• [x] I agree to respect and uphold the Cucumber Community Code of Conduct
• [ ] I've changed the behaviour of the code
  • [ ] I have added/updated tests to cover my changes.
• [ ] My change requires a change to the documentation.
  • [ ] I have updated the documentation accordingly.
• [x] Users should know about my change
  • [ ] I have added an entry to the "Unreleased" section of the CHANGELOG, linking to this pull request.

---

This text was originally generated from a template, then edited by hand. You can modify the template here.

[coveralls]

Coverage: 98.561%. Remained the same when pulling 7c7ddae325fd5e43cd9ec181d7385a5947b322ab on chore/deprecations-housekeeping into 90ebb805a8c1afc0d2118e6edd70343e13fc7c66 on main.

[mpkorstanje]

From a process point of view I have only one remark:

except maybe SummaryFormatter which is the one people would reasonably extend when building their own.

If this class is open for extension it should be documented as such. And be tested against the specific use cases it aims to support.

Otherwise it will be hard to evolve the code in the future.