Remove BoxConstraints::min

[PoignardAzur]

I'll probably break this into multiple PRs at some point. This creates a lot of snapshot test errors.

[jaredoconnell]

This is what happened to the calc example:

With min removed, we're going to need to add the RequestedSize enum to the layout calls. That will allow us to account for the case where min is currently used. Basically just tell it to fill the space instead of use the min size.

When I get a chance I'll experiment with that route and if I come up with something useful I'll submit a PR.

[PoignardAzur]

I'd rather have an "expand" attribute on the button itself, if we go that route. I still don't like the idea of telling a widget "you need to be at least / exactly this size".

But yeah, these examples need to get un-broken before this can get out of draft.

[PoignardAzur]

For the record, I think the way to go is to split this PR into multiple chunks. So first we might call loosen() in a bunch of places where we currently pass tight constraints, and see which errors pop up. Then once roughly all constraints we pass are loose, we can remove the mix parameter without disruption.

[jaredoconnell]

I'm currently experimenting with replacing BoxConstraints itself.