Nice demo (it assumes target exists though, which is not a safe-assumption) ... my quick thoughts around this issue/question:

1. display is a general-purpose (imho) last-resort utility to show something on the page when you don't know any better
2. being general purpose, it would be weird to have tons of checks out of possibly unknown imported libraries out there, it will fail in a way or another with one library or another
3. you do know what you are doing, and you don't need current_target() to accomplish that ... you just target that element created to show your animated demo and call it a day, the usage of current_target() out of a tagName check and whatnot, when you want to just inject whatever to_jshtml() returns to a DOM element, feels like way easier, and under your control, than us best-guessing your intent out of a more advanced use case that has specific requirements
4. because of the previous point, your whole demo could be simplified as such: https://pyscript.com/@agiammarchi/attempt-at-matplotlib-animations-copy/latest?files=main.py

Accordingly, I don't think there's any action we should take, but happy to hear otherwise from @ntoll or others.

Hang on a minute, I just noticed that matplotlib animations actually implement _repr_html_ so this might be a precedence issue. Why does savefig appear before _repr_html_ in _MIME_METHODS?

Hi folks,

Just catching up here... very cool demo. I think we should at the very least document how to do Matplotlib animations in PyScript along with a working demo. We can do this once we're all agreed on the best recipe to recommend. Likewise @jburgy - any feedback on code, or a PR to change order of calling (e.g. savefig and _repr_html_) would be welcome. I didn't write that code, but I don't think the original author was considering animations when they wrote it.