RFC: ignore vs. xfail

[brson]

Related: #11490

We use two different words in our various testing tools to mean 'this test is broken': 'ignore' and 'xfail'. I would dearly like to standardize on a single one. My inclination is 'ignore' and we've already started moving in that direction, but I have one concern about this: as we expand our testing to more documentation 'ignore' could potentially mean many things, not just 'ignore this test', so tagging snippets with '~~~ignore' is potentially confusing.

[jfager]

In Javaland, junit gives you @ignore("reason").

[brson]

NUnit also does ignore (inherited from JUnit presumably).

[sfackler]

I have no idea what xfail is supposed to signify. Ignore seems much more straightforward.

[alexcrichton]

Closing, we're using ignore everywhere now.