feat: Add a feature that allows piet to be used in no_std environments

[notgull]

As below, so above.

The main hiccup here is that I can't use Box<dyn std::error::Error> on no_std, so I just make it so the BackendError variant of Error isn't there on no_std. There is probably a more elegant solution to this problem.

[xStrom]

I haven't looked at the changes in any depth, but the CI failures should be resolved.

[notgull]

I can't replicate the failures locally

[xStrom]

Is this still something you would like to see merged?

If so, then the current problem is that the CI seems to be stuck in 'waiting'. I think the best approach here is for you to rebase on master and force push to your branch to re-trigger the CI for this PR. Thanks!

[notgull]

I still want this, it's just that I've been busy with more pressing issues.

[notgull]

I have no idea why the piet-core-graphics implementation is failing

[ratmice]

@notgull #562 has a fix for the piet-coregraphics failures, but also bumps some dependencies.

I got that other failure once about local connections, but another time it went green so I'm not sure if that check is spurious or why exactly it finally went green for me.

[xStrom]

Alright, some significant time has passed now. The macOS dependencies have been updated and we have a new ARM based macOS runner for CI. The master branch is passing CI nicely right now.

It would be nice to get this landed. Please rebase on master and let's see if we can get the CI to pass for this PR. 🤞