Off the Beat: Bruce Byfield's Blog

I'm a confirmed KDE user. I have seven desktop environments installed virtually or on my workstation's hard drive, but I spend most of my time in KDE. However, in the last year, a strange thing happened: I started using GNOME more.

This was the last change that I expected. Ever since GNOME 3.0 was released, I've complained about the overview mode, which seems better suited to a mobile device than a workstation or a laptop. I've complained, too, about how it restricted users by such features as automatic management of virtual desktops. Most of all, I've complained about GNOME's slowness to respond to user criticism, or even to acknowledge it, and the defensiveness of its designers.

I still consider these criticisms generally valid for the time when I made them. However, about a year ago, GNOME made a decision that has yet to receive the credit it deserves: It decided to encourage the use of GNOME Shell extensions. Officially, it supported a core set for classic GNOME, but, unofficially this course also promoted the other extensions as well. With this decision, I quickly started to look at GNOME from a different perspective.

The Joys of Extensions
Rightly or wrongly, I consider encouraging the use of extensions a graceful response to the criticism showered upon GNOME. Without any GNOME developers losing face, it provided much of what the critics had demanded -- in particular, a GNOME 2 experience. The policy should have come earlier, but at least it came.

More importantly, encouraging extensions removed the most important complaints about GNOME Shell. For those of us who dislike the overview mode, extensions allow us to ignore it altogether, using a single screen and seizing control of virtual desktops for ourselves. For those us who complained about a lack of customization, extensions suddenly gave us choices. Suddenly, the complaints became invalid. Even those who liked GNOME 3 found themselves with more choices than many other desktop environments could boast.

Admittedly, some extensions conflict with others. But these conflicts are relative rare, and frequently temporary.  Nor do most of them matter, since key elements like alternative menus and panels are usually available from more than one extension. If one extension fails you, frequently you can find another one. This is exactly the kind of diversity that as a long time Linux user I want on my desktop.

Back to basics
Encouraging extensions also had another consequence -- one that I'm not sure that those involved fully appreciate or would accept. Early in the GNOME 3.0 release series, GNOME developers seemed concerned with the purity of their vision. It was not GNOME that should adapt to users. Rather, users should adapt to GNOME's conception of a desktop.

However, once GNOME decided that users could customize with extensions, this position became untenable. GNOME was adapting to users, so debating whether it should became a lost cause.

Instead, designers turned their attention from the big picture to look at the details. After several releases in the series, this move was due to happen anyway, but supporting extensions probably reinforced it.

But, whatever the cause, the result was to re-focus GNOME on its strengths. Historically, GNOME has always been more successful with incremental changes, developing a distinctive minimalist design that has sometimes been controversial, but for the most part has shown a concern with usability that few other desktops can match.

Personally, I prefer KDE's completist approach, although it sometimes leads to  disorganization in the menu. But I have always admired GNOME's minimalism, even though I am not always in the mood for it. GNOME's minimalism is the first and longest-lasting attempt on the Linux desktop to think about design and usability, existing before Ubuntu's Unity was even imagined, and I'm glad to see that the accumulated expertise has not been thrown away.

Just as importantly, GNOME's current attempts to re-design dialogs and utilities to fit GNOME 3 are one of the few efforts by a major desktop to think seriously about design today.  

Unity's efforts seem largely unsuited to the desktop of anything larger than a phone, and appear to have strayed into dubious byways such as adding ads to the dash. By contrast, GNOME is community-oriented design by people whose only concern is to improve the design. I sometimes drop by the blogs of designers like Richard Hughes and Allan Day, and while we have clashed in the past, I have learned to appreciate their focus and hard work as I learned more about it. While I may not always agree with the design team's priorities, it is one of the few teams in free software thinking about design at a time when visible innovation is rare on the desktop. As they slowly transform GNOME and give it a consistent look, their work has -- entirely unexpectedly -- become one more reason to appreciate GNOME.

Taking another look
But why take my word? Today, GNOME 3.10 is being released, and you can see for yourself. One of the drawbacks of extensions is that what you get when you install GNOME often differs depending on the distribution, but in any variation, GNOME has evolved to become far more promising than GNOME 3.0 might have predicted.
Of course, users have long memories and can be slow to forgive. No doubt some have happily settled on Mate, Xfce, or one of the other half dozen alternatives they have explored since abandoning GNOME.
Still, take the time to install and tweak GNOME to your liking. You might discover that it is no longer the desktop you thought it was.