Repository navigation
The state of state in layouts #1869
ianyh
announced in
Announcements
Replies: 1 comment
|
Why not offer a "Classic" fork for anyone that wants to keep the existing state functionality, and then push main forward without state as desired? It might not be the best experience for those that need to re-download the stateful fork — but it also lets this project progress with your continued developement. |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
I designed myself in to a bit of a corner with state tracking layouts. I have been experimenting with handling native tabs, and I seem to be bouncing between correctly deriving state at any given time or tracking the change events precisely. The architecture doesn’t lend itself well to the latter, and my solutions are getting increasingly convoluted.
Given the complexity of building the state tracking on top of current architecture, I am considering dropping stateful layouts. This doesn’t end support for state entirely. Rather, it would make state solely user driven instead of also system event driven. This would end support for some custom layouts that are relying on system events, as well as the native bsp layout. I’m not super concerned by the latter, as there are many other window managers that have much more robust bsp support, but the former does concern me a bit. I would rather not break existing custom layouts, but they also might already be broken because they can drop or receive inconsistent events.
Generally speaking, I think dropping the event tracking would significantly reduce the scope of the project, which would also significantly increase my capacity for maintaining it, but I’m torn on whether it is the right move or not. I don’t know whether anyone has ideas or opinions. I think I just needed to get the thought out there before I actually do anything.
All reactions