BTW, there are a lot of files that get touched when I run black. What is the policy related to formatting and linting?

Ah I think we forgot to run black on the examples, they're only run on the source files I think.

We haven't made a strict linting policy yet but we should. I'll see if I can export my flake8 options from pycharm.

For this example, I think it'd be useful to add a note that if you want sliders to scroll through one or two dimensions you should use the ImageWidget in either Qt or jupyter. Since we don't have within canvas UI right now.

For this example, I think it'd be useful to add a note that if you want sliders to scroll through one or two dimensions you should use the ImageWidget in either Qt or jupyter. Since we don't have within canvas UI right now.

👍

Odd though, because I have PySide6 installed, and it selects glfw. Will have a look why this is.

Oh, it only looks for PyQt6 for now. Ref #384

Oh, it only looks for PyQt6 for now. Ref #384

Yup! I updated that issue with the specifics.

Backend selection is a bit iffy. Will still select glfw because wgpu prefers it over qt. Something for another PR :)

Yup I think that's why our Qt example explicitly chooses the canvas

If this is good to go you can mark it as ready for review to trigger the CI to run, and then copy the new screenshot from the regenerate action to the corresponding examples dir.

CI: AttributeError: module 'fastplotlib' has no attribute 'widgets'

But when I import fastplotlib as flp I can use flp.widgets. Which behavior is to be expected? :)

ImageWidget is intended to be used from top level, i.e. fpl.ImageWidget. Not entirely sure how this is happening.

This is. Perhaps at some point the ImageWidget relied on ipywidgets?

This is. Perhaps at some point the ImageWidget relied on ipywidgets?

Ah yes good catch! That shouldn't be there.