tree: skip empty inactive workspaces

[kotarac]

Handles point 2 from #519.

Not sure how it should behave depending on ws.attention_requests. Can those happen if the workspace is empty?

[mahkoh]

The current behavior is intentional so this will have to be a configuration option. Maybe something with five variants:

• preserve: does not do anything for empty workspaces
• destroy-on-leave: destroys an empty workspace when it is switched away from - this is the current behavior
• hide-on-leave: hides an empty workspace when it is switched away from
• destroy: destroy an empty workspace when it is not active
• hide: hide an empty workspace when it is not active

Hide differs from destroy in that reopening the workspace opens it on the same output it was previously on.

Can those happen if the workspace is empty?

No.

[kotarac]

Sure, let me see if I can get that done.

While we're on that, I already had a naive solution that sorts workspaces (point 1. from #519).

Would you prefer that configurable as well?

[mahkoh]

That's more complicated. You can manually drag workspaces around so you'd have to think about how these two things interact.

[kotarac]

That's more complicated. You can manually drag workspaces around so you'd have to think about how these two things interact.

I was thinking making it configurable and if sorting is enabled having dragging disabled.

Hide differs from destroy in that reopening the workspace opens it on the same output it was previously on.

Is hide a new concept or is there precedent in jay?

[mahkoh]

I was thinking making it configurable and if sorting is enabled having dragging disabled.

That sounds reasonable.

Is hide a new concept or is there precedent in jay?

It's what your code does at the moment. Otherwise it is new.

[kotarac]

Closing this for now. Will take a stab at it according to #541 (comment) sometime.