The GNOME project at 15

I am so glad you mentioned this, because it epitomises what is so wrong with Gnome 3 (and I am also glad you find it usable - at least some do).

Like many other desktop features, workspaces have been born out functional necessity: how to have a larger desktop with the screen you have been given. So, workspaces (or in the olden days viewports) were fashioned to represent a "zoomed in" view into a larger desktop. This necessitated the existence of a visual workspace switcher, to tell users where they were and where they could go next and about what was on other ports/spaces. It was easy to visually identify all this information, immediately. Some folks arranged their workspaces into a row. Some into a grid etc. This was a concept, born out of functional necessity. Dynamic workspaces are a visual hack, just for the sake of it. They are the equivalent of rearranging your windows when one of them is closed. No other GUI does that.

Similar with overview. On a smartphone, an overview is a functional necessity, because there isn't enough screen real estate to present both the applications menu and the currently running application(s) on the screen. Hence the home screens (which is what overview is). And, of course, once again, Gnome 3 follow down the wrong path on devices that have no such functional need (desktops, laptops).

Minimised (or in the olden days iconified) windows came out of another necessity. On a multitasking system, one often has more than one thing going on at one time. And, occasionally, there are so many that some of them have to be taken out of visual range, so that more important tasks can been seen. This then necessitated the invention of iconification. This in itself was not good enough, because iconified windows would get lost behind others. So, a taskbar was invented, which was then always visible, so that users could bring back tasks they were working on easily. Gnome 3 does away with this, but it leaves user an option to still minimise windows (in a surprising about face when compared to the rest of the system, through a configurable option). Of course, the windows then go into oblivion and cannot be brought back unless overview is entered, where they miraculously appear as normal windows. Another concept broken.

Only the most obnoxious of web sites will try to maximise your browser window. So, good browsers have an explicit option that stops such obnoxious web sites in their tracks. On the other hand, Gnome 3 window manager will now do that for you. It will maximise you windows when it sees fit. The rationale? "Displaying multiple windows at the same time means that screen space isn't used efficiently, and it means that you don’t get a focused view of what it is that you are interested in." In fact, the exact opposite is true. User should decide what is and isn't the most efficient layout of windows. Contrary to this new Gnome 3 philosophy, some people can and do more than one thing at once.

I could probably write more examples (and I have here on LWN, in many comments), but I think the above already sufficiently illustrates what I (and many others) have been trying to tell Gnome developers for months now.

So, when you say:

> It's almost as if the usability experts know what they're talking about.

I have to disagree in the case of Gnome usability experts. Strongly. Based on the above objective complaints.

Havoc can wax lyrical about Gnome 3 team and I understand he knows many of them. But, from what I see, things are not going in the right direction in many of the areas. Long standing functional concepts have been broken. And nobody seems to be listening.