Feature Request: Support for one-way control

[b0o]

Really awesome project! One thing that would be nice is that I want to control my Mac from my Linux computer, but not the other way around. When I am focused on my Linux computer, I don't want my Mac's trackpad to be routed to my Linux computer. I always want the trackpad to control my Mac.

Maybe something like this in the config:

[right]
hostname = "my_mac"
activate_on_startup = true
client_only = true

(Note: I am currently using #131 on my Mac)

[feschber]

This is a good idea!
I have already been considering this, since it would open up a few more usecases like multiple devices controlling one other device.
My plan would be to have a separation between "incoming" and "outgoing" connections.

[Dregu]

My use case for one-way control would be that the laptop I'm using to remote control my desktop is simply always behind my chair on the couch, so being able to accidentally remote to a laptop I can't see is just detrimental.

That said, what I would actually like to do is keep two-way control as an option for special occasions (when it's actually on my desk), but simply disable any kind of mouse-on-screen-edge detection and instead use keyboard shortcut to connect to a specific host (or release if already connected).

This really goes hand in hand with the other two suggestions, but after disabling screen-edge and adding keyboard shortcuts, hosts should be able to just exist in the network with a name like [steve] instead of [right] or [left].

Currently I am launching lan-mouse with a keyboard shortcut on demand on the laptop when I feel like poking something on the desktop/TV, so that I wouldn't accidentally remote to the laptop, but there's still the extra step of dragging the mouse to the edge.

So yeah, just some thoughts how having these three options instead of a simple one-way switch would be a lot more versatile, while also achieving one-way functionality if configured that way.

[feschber]

So this is not fully what you are describing yet, however the first and most difficult step is there: Device A controlling device B is now independent of device B controlling device A.

What you are describing is somewhat the same as #26.

So my plan now is to

a) make automatic cursor release on the remote device optional (this would also be a workaround for #132)
b) implement an option to add a keybind for toggling between devices

Removing the capture activation barriers on the sending side is somewhat more difficult because of the way libei works but it might be doable as well.