DeprecationWarnings in 0.3.0 tests

[ChanceNCounter]

NOTE TO CONTRIBUTORS: This only applies to DeprecationWarnings, and only in test files. Any other warnings, as well as DeprecationWarnings in code, are bugs!

---

The tests for pending version 0.3.0 now raise (as of this date) 12 DeprecationWarnings. These are annoying, but pytest will only print one of them, and I'd rather not suppress them.

My theory is that if we annoy ourselves for the next 6-12 months, we're guaranteed to remove the deprecated behavior the very instant it seems reasonable to do so.

If nobody objects to this, I'd also like to pin the issue so that people don't spend the intervening months reporting it, and submitting PRs to fix it =P

[ChanceNCounter]

(rfc @krisgesling @forslund @JarbasAl)

[krisgesling]

Are these the lang=None warnings? Given they're intentional, should we assert that the warning is raised instead?

My concern with leaving known warnings is that people then stop paying attention to new legitimate warnings that shouldn't be ignored.

[forslund]

I do think keeping the warnings is a good idea.

Testing for logs is generally not a good idea in my opinion, very fragile and doesn't actually test the functionality.

The warnings should however state in addition to it being deprecated something like "this will stop working completely in version 0.4.0" so it's clear that it's a planned change.

[JarbasAl]

i'm with ake on this one, deprecation warning should be there with that extra note

and this issue should be pinned so it keeps support requests under control