What are your thoughts on the upcoming COSMIC DE?
What are your thoughts on the upcoming COSMIC DE?
Basically title, with System76 moving from gnome to their new rust built COSMIC environment what are your thoughts?
I'm incredibly interested in COSMIC DE! For multiple different reasons, actually.
- Rust - I'm very interested to see how performant/memory-efficient this DE will be compared with other DEs. Also, I wonder how the Iced toolkit will evolve and be adopted in other projects.
- Benefits over GNOME - I'm looking forward to seeing how much out-of-box customizability and features come with COSMIC over GNOME (which I'm currently using).
- Maintainability going forward - Since the DE basically started from scratch and is using a much better language for robust software, I wonder how much easier and faster it would be to maintain the desktop environment. This potential improved maintainability could be huge in overtaking other DEs sometime soon.
A couple of months ago our company decided to standardise on only one GNU/Linux distro and they chose PopOS. While the default desktop is better then stocj GNOME it was still far away to what I am used from the powerful, featureful and customizable KDE Plasma so after about two weeks I switched to KDE Plasma (unfortunately they have an extremely old version in their repos, but still much better).
I can only guess that Cosmic will be on pair to their current improved GNOME but will still be way lacking compared to what even an old KDE Plasma offers. And I would also much more like to see if they put more attention to keeping more updated KDE Plasma and KDE software packages in their repo. Even for Cosmic I think they would be much better of basing it on the extremely flexible and configurable KDE Plasma base and make it a heavy modification of this.
There is a community KDE spin in the works, with a more up to date plasma experience. https://fosstodon.org/@pop_os_kde_spin
Personally, I would have preferred if they would have embraced KDE instead and dedicated resources to improve KDE even further. KDE is pretty darn good already, especially over the last year or so with all the bugfixes and feature improvements.
In any case I am happy System76 is moving away from Gnome at least, and coming up with a full-featured DE of their own rather than having to maintain hacky extensions on top of Gnome to make it usable. As a bonus, the Iced toolkit should see improvements and perhaps become the goto toolkit for Rust GUI development.
I welcome new and innovative DEs. I am tired of how some organizations handle things. I must however, acknowledge that with every additional major DE comes further fragmentation. COSMIC seems to really care and for that reason I am rooting for them. The negative effects, however, are always in the back of my mind.
Skeptical. Writing a graphical UI toolkit is a freight train of work. I'm positively curious about anything that's not GTK but I'm not sure going with a new toolkit is the right decision. Qt is the mature kid on the block that's been proven in more environments than I can count. Moreover it's a complete application framework with a ton of convenience libraries needed for speedy development already included. I guess those can be supplanted in the form of separate Rust libs. Personally I'd have gone with Qt for such a project but I'd be happy to be proven wrong.
This is coming from personal experience/opinion, but after trying to create a simple app in GTK4 Rust bindings I was so confused because of how alien the programming style was compared to typical Rust programming. After trying Iced it was much simpler and made so much more sense, no silly decorators or anything, you can define the view and the update loop separately, and interactions are handled by messages using pattern matching. The inheritance based OOP doesn't work well with Rust, and Iced has none of it, because it was made for Rust specifically.
I'm guessing QT bindings are similarly in a different style of programming and can't imagine that meshing well with native Rust code. Iced has a lot of merits to it and having the opportunity to both help it develop and use a native Rust framework in a Rust project makes a lot of sense.