GNOME Shell + Mutter 3.37.3 Are Out Roaring With Better Performance

[bdog]

Now you just listed every single DE from the top of your head for trolling purposes. User experience on KDE is really, really bad. Especially the desktop effects are in the same sad state that they were when KDE 4.1 was released. Everything blinks and twitches and is erratic. Plasmoids suck ass and do not work well. Most plasmoids are complete crap in the "icon mode" when thrown on to the taskbar. Most plasmoids, actually, are useless crap even in regular mode.

(I was a KDE user from 4.1 through 4.5 and then switched to Gnome for good.)

P.S. I've tried using some Deepin apps but they don't even reach the same quality as my abandoned pre-alpha state hobby projects.

True True! GNOME has the best vision. As for strategy: desktop Linux needs to band together for the time being and produce a world-class desktop experience. GNOME is best positioned to do this. Once GNOME fights off Apple and MS we can focus on polishing up the other desktops. We need to think strategically to beat our common enemies.

[brent]

Sounds like Gnome Icon Grid is the new Crysis... 30..40 fps for a basic OS-Feature. At the same time when the refresh rate of monitors finally is over 60 Hz.

You forgot that this was running at 4K resolution on an older Intel GPU! You can't really expect smooth 60 Hz on such a setup.

[DanL]

Sore because some KDE user decided to troll a bit and say something stupid?

LMAO at you accusing people of trolling and saying stupid things. Those who live in glass houses...

[Vistaus]

I want support for blurred opacity.
Also want the animation to be bottom-up sliding (like iOS and Android) instead of diagonally folding.
It is great that there is ArcMenu and Dash-to-panel which makes GNOME better.

You can keep repeating that in every GNOME thread on Phoronix (seriously, I haven't seen a single GNOME thread where you didn't mention this!), but that's not really going to help.

[Vistaus]

If you question shell UI/UX, there are themes and extensions. Gnome 3 was made scriptable and hackable on purpose as its creators knew that their UX will be radical and not everyone will like it. Whatever you like, Gnome replicate it be it.
If you question the userland apps, thats more tricky. Personally I was once like you but became a changed person after one week of trying it with a open mind. If you still don't like the core apps, you are free to replace them with whatever you like.

I think it is interesting that you include OSX in your list, a OS with a not radical but horrible UX and broken by design core apps.

I'm pretty sure I've heard time and time again on Phoronix that GNOME Shell never meant to support extensions, so I doubt your statement is true.

Also, even if it were true, the same can be said for other desktops. KDE Plasma, Deepin, Cinnamon etc. all support scripts and extensions by default (KDE even more so than any desktop out there as you can use GHNS in pretty much every part of it, not to mention different theme engines like Kvantum).

[Vistaus]

P.S. I've tried using some Deepin apps but they don't even reach the same quality as my abandoned pre-alpha state hobby projects.

Deepin 15.11? Yes, not the best quality apps (though still better than GNOME IMHO). But they improved a whole lot with Deepin V20 - the quality is very good now (and they also added a few new core apps, as well as a few WIP ones that aren't released yet, like a new mail client for example).

[Mez']

Deepin is like TikTok. Byebye security and privacy.

Gnome is like Mac Os. Bye bye user possibilities to make decisions on their own. Adios user feedback, diversity of use cases and workflows.
Hello dictatorship of the one-track thinking. Hello ostrich developers, and heads buried in the sand.
It's really good for followers, sheeps and trivial workflows. Not so good for leaders, for people who make use of their critical mind, who are ahead of others and go the extra mile.

Beside performance improvements, I don't see anything in 3.38 that is aimed at solving this. They still assume their users are dumbwits with no will of their own. And boy, many seem to be.

I really hope Budgie 11 (whenever it comes) will settle this once and for all (for me), as the focus is much more user-centric than developer-centric as in Gnome.

[treba]

I want support for blurred opacity.
...

You may want to have an eye on https://gitlab.gnome.org/GNOME/gnome..._requests/1339 and friends. Things are in progress, but some heavy lifting needs to be done to get there.

[Mez']

Mez' While you are waiting for Budgie11 please think about why all the big distributors just went with GNOME and added a few extensions and theme changes as needed. Clearly they don’t share your opinion.

If they added so many extensions and changed the default theme, they clearly share my opinion...
Most distros using Gnome just stick to mainstream DEs to be safe and because of the ecosystem around it.

Also, if I'm talking about Budgie 11, it's because 10.5.1 is like wayland, it's almost there but not quite yet. It's not as stable as Gnome, lacks a few use cases on the multi-monitor experience in particular and the ecosystem is not as wide (although it's GTK and benefits from it). It's still behind Gnome as it doesn't have the same longevity as Gnome, but it's seriously catching up.
I'm really looking forward to the next steps, as it fits a much broader span of workflows than Gnome, and it really make you the decision-maker which means you can tweak in very different directions.

[You-]

OTOH the recent leaps in performance happened after Canonical switched back to Gnome.

Just to address this point: correlation is not causation.

Other things happened around the same time, such as Gjs getting updated and was able to keep up with the latest stable MojJS releases. This work took multiple releases. On top of this infrastructure work including the later sysprof support that was carried out allowing both profiling and further performance work.

These things enabled Canonical employees and others to find bottlenecks and carry out the performance work.

Without work on other parts, the performance work could not have been done.