Antony Lee

Thanks, this looks very nice indeed. Perhaps you can e.g. make a PR to modify the `dataframe.py` example to use this style (and perhaps use more interesting fake data, too)?

That's what the warning says: no one wrote a picker for QuadMeshes. Someone needs to write a compute_pick(), a get_ann_text() and possibly a move() implementation for that class.

Are you running into this issue for a real use case? The main problem I am aware of is with using figimage() (so "per-pixel data") with pdf output (where the...

No need to apologize when you're proposing a fix :) (though it does look like the PR CI is failing -- let's continue the discussion there)

https://github.com/matplotlib/matplotlib/issues/16268#issuecomment-576412375 may be related.

I don't have the bandwidth to add it right now, but this seems like a reasonable request.

No, this is the way to do it. You can consider making the dots transparent with `c="none"`, though. A good improvement would be to add a method to "programmatically add...

That's not possible right now, the point is that this would require quite a bit of work to enable (not necessarily *very* hard, just stuff that needs to be done...).

Strange, this is what I get locally: ``` $ python run-mpl-test-suite.py --pyargs matplotlib.tests.test_backends_interactive -vsk auto_backend_selection Test session starts (platform: linux, Python 3.10.5, pytest 7.1.2, pytest-sugar 0.9.4) cachedir: .pytest_cache hypothesis profile...

It was with an editable install, but I have now tried with a normal install of mplcairo and matplotlib (in a clean venv) and everything works fine too..