morphorm VS bevy_ggrs

I started working Rustycan - a small UI framework that focuses on developer ergonomics and uses other frameworks (such as egui/druid/html) for the rendering, except for the layout which I'm planning to use morphorm.

That said, I'd be interested in migrating towards something closer to the Hierarchy trait that morphorm has pioneered. If this is something that's interesting and important to you could you make an issue?

jkelleyrtp has been experimenting with other UI layout strategies, including SwiftUI. On my end, I really want to see morphorm support (based on subform): initial experiments are promising, and it's "refreshingly simple".

There's a few critical subtasks here: - determine the data flow model we'd like to use for our UI. We'd like to integrate tightly into the ECS, but need to figure out how to reduce the boilerplate and improve reliability around working with hierarchies. - swap our layout library. Our current dependency stretch implements the flexbox algorithm, but is unmaintained and has critical bugs :( The three main options here are to fork flex, move to the new morphorm crate, or write a layout library from scratch - consider rearchitecting our UI to be more flexible and compositional: splitting the massive Style component into several parts and moving to a "UI is a collection of behaviors" paradigm - build out more widgets! - more docs and examples!

As of today, I believe multiplayer support is not in the short term roadmap. There are community plugins like bevy_ggrs that could be used for that.

You could create a custom stage (by implementing the Stage trait) that stores its own schedule and then use schedule.run_once(world) to run all systems in the schedule exactly once. I did this in bevy_ggrs , a bevy plugin for ggrs, a rollback netcode library.

Regarding rollback networking, there is bevy_ggrs, which I think works really well.