Wow, this is great stuff! I'd love to try this, but I get that it's still a bit away from being a usable prototype. I'll definitely keep my eyes peeled for this.
this post was submitted on 26 Jul 2023
29 points (100.0% liked)
Gnome
34 readers
1 users here now
The GNOME Project is a free and open source desktop and computing platform for open platforms like Linux that strives to be an easy and elegant way to use your computer. GNOME software is developed openly and ethically by both individual contributors and corporate partners, and is distributed under the GNU General Public License.
founded 1 year ago
MODERATORS
I just love that on the Linux desktop this kind of innovation is possible. They are paving the way for an entirely new desktop experience and I'm so excited to see where this goes!
Hmm, the Mosaic concept is quite interesting, but I feel like, personally, I wouldn't want to use it over traditional tiling.
For example, my workflow does involve moving windows to a new workspace to have them maximized, but I do that very deliberately. I want control over which workspaces are next to each other, so I can quickly switch between related applications.
But I am usually on a smaller screen, so:
- I'm not really affected by the overly large windows, they're trying to resolve,
- I need lots of workspaces and that predictable navigation between them to fit all the windows, and
- I would encounter the screen-full edge cases with Mosaic layouting quite regularly.
Then again, they do have the idea here that if the screen is full, it will basically switch to traditional tiling. If on small screens, the screen is pretty much always full, and if the traditional tiling works well, then all my objections would be void. 🙃
Loved it
The biggest reason for me being excited about System76's new DE is that they'll implement 1st class tiling support. Tiling being only possible through extensions on Gnome comes with limitations/bugs I don't experience on tiling compositors like sway and hyprland.
That's why I'm really happy to see work going into native tiling on Gnome and I'll definitely give it a go once some working prototype is available.
Yea, this would really be a step forward, but also a step out of the comfort zone of some newcomers. We haven't seen much but I hope it won’t be to aggressive so problems like described in the article, where windows get to big, or to small wont happen.
Yes. The biggest hurdle is application support. Gnome apps might be quite convergent and might even support this new "max size" API but most cross-platform apps probably won't.
So hopefully they'll also develop ways for this to work well for apps that don't provide any info.
Kinda missing any word on corner cases like multihead setups and fake "fullscreen" (e.g. for braindead games having ideas about screen layouts not offering full sized display resolutions). Both are still a pain in the neck especially when combined. The current system works okay-ish (with gamescope solving one of the worst headaches, no more fiddling with virtual displays that extend over all real displays).
For coding e.g. I often wish vscode to span over several displays. I'd start jumping in circles in anger if something would mess with a carefully aligned size and tabs inside of it automatically ruining the initial "sorting work".