Personally, 2019 was finally the year of the Linux Desktop for me. Proton from Valve has totally changed everything for me. It runs AAA titles such as The Witcher 3, Grand Theft Auto V and even Final Fantasy XIV (with a little tweaking) out of the box at 60 FPS! That was my final holdout and it freed me up to just run Manjaro outright :)
Although I've been running Linux on my desktop for a long time, I indeed also kept a Windows install around. I've tried (and succeeded) in using Wine over the years, but it can be tricky to setup, and known good setups from e.g. 2004 at some point stopped working and I kind of gave up on it.
Proton changed all that: zero config compatibility for many of my Windows games, even demanding ones like you mention. Simply fantastic. December 31 2018 I removed my Windows partition so in that sense, also for me 2019 is the year of the Linux Desktop.
While everyone around me is rejoicing, I am somewhat sad that a single commercial entity managed to do in a year what the whole FOSS community could not do in twenty.
They were standing on the shoulders of giants (Wine), which is like 90% of the work, but it seems you still need to involve commercial interests to put the necessary polish to make the product really useful.
An important ingredient to Proton in DXVK, which was a new project and the author of which was quickly funded by Valve as it greatly improved performance, especially AAA+++ titles. Secondly, Valve funds Crossover, the commercial entity already responsible for large parts of Wine. The heart of Proton is thus free and libre software.
What sets Proton apart from alternatives like Lutris, which also configure Wine, DXVK, and other components, for you, is that it is zero conf, integrated in a tool (Steam) you probably are already using if you play games. Personally, Proton is the only tool where the autoconfiguration actually always works.
Well this grumpy old man switched away from gnome when they went 3.0. I’d say I don’t know anyone who stayed. Some of us went to xfce and others tried out mint which ships something more like 2.
Perhaps the Gnome desktop. As a big fan of the 2 series I don't know what the goal was with the 3 series. More dumbed down than mobile OSes... I took refuge with the KDE project, which, with Plasma 5, really got into its own, but Mate is a good alternative as well.
Gnome, I know it's the default on most distros, but I can only imagine that is because of inertia. Who ordered what Gnome 3 offers?
But more seriously, this is a silly question for two reasons:
1. There's no such thing as 'The Linux Desktop' unless you create a strong tie-in between the kernel and the desktop environment. IIANM Gnome is kind of doing that through systemd, although I might be wrong. Anyway, other than that, it's "X- and Wayland-based desktops"
2. Are there libre, non-X, non-Wayland desktops which are doing better than KDE, Gnome, XFCE, Mate, Cinnamon and the rest? ... I don't think so. Maybe on mobiles.
> Isn't there potential to update the desktop into the 2020 version of itself?
(shrug) Whatever...
> And isn't it possible for GNOME to be the organization that does that?
Sounds like they want to remove even _more_ of the UI, so that in their dialogs and windows you really can't configure anything at all. Welcome to the 2020 version of lameness.
I would say so. You can’t even launch a vanilla Linux instance on AWS and connect directly to its desktop, at least without going through lots of hoops.
I blame the KDE vs Gnome fragmentation, and over-customisation.
There's a ton of fragmentation on Windows. There are many more toolkits and paradigm's on that platform.
And regarding fragmentation, GTK+ vs. Qt has never been the main issue. It's been different distributions and the fact that app creators cannot ship apps easily to all distro's themselves.
But, I believe, the main issue behind Linux Desktop has always been lack of corporate interest. There was some in early 2000's but never anything serious.
Windows and Mac "are" ecosystems due to financial motivations.
> I blame the KDE vs Gnome fragmentation, and over-customisation.
Fragmentation is one of the double-edged swords of all open-source projects. It leads to greater innovation, at the expense of duplication and wasted efforts.
I would still rather have the freedom to choose and contribute to a fragmented ecosystem, rather than have to be forced to deal with issues in a black-box propitiatory OS with "unclear" motivations...
