GNOME Is Losing Relevance On The Linux Desktop

[spirtbrat]

Michael is not the one saying Gnome is loosing relevancy, the Gnome developer who wrote the blogpost is. Michael just happens to agree.

Of course he didn't wrote the "staring into the abyss" article. But he thought it is good opportunity to point out the Cinnamon and the Mate projects, as they are bringing those good old memories about the GNOME 2.0 desktop.

There are more of his thoughts, likings and opinions. Like those for example:

"The negativity towards the GNOME project isn't surprising since the botched GNOME 3.0 release."

"Prior to reading Otte's GNOME abyss blog post last night, I myself was thinking of what desktop environment to use next. For nearly two years my main production desktop has been Ubuntu 10.10 with GNOME 2.32 virtualized within Mac OS X on an Apple MacBook Pro. With my new retina MacBook Pro, it's time to move past Ubuntu 10.10 and the most pressing problem has been missing GNOME2."


Every interactive application - being either console or graphical one - is enforcing a certain way of work flow. A way that the developers thought it may be the best one for this king of activity. Take for example the once hugely vibrant competition between Vim and Emacs - two overpowered text editors with very different manner of achieving the same goals.
Also the various console mail clients.
Or even the numerous Unix (-like) kernels at the time when they were not so different from one another.
The thing that made them distinguishable were their philosophy which explains their different way of achieving same common goals.

GNOME 3 _is_ different and it is different for a reason. You _cannot_ work the same way, like you did with every graphical interface since Win95. If you need to use your old work flow for whatever reason (no time to figure out how to be as productive with the new tool, no appreciation of the benefits it brings you, or just plain stubbornness), then you _have to_ chose another graphical environment. Here is a recent interview that Michael could've appreciated if he wasn't so GNOME 3 disappointed and not following any GNOME-related news source:

Treat Gnome3 as something new” by Allan Day

And to address the "disappearing GNOME developers" claim, I will highlight another 2 short weekly reports:

1778 commits, in 207 projects, by 207 contributors in a week

1816 commits, in 178 projects, by 214 contributors for a week

both of which are from this month.

[blackiwid]

omg you postet your comment 4 times? lol ^^

I agree mostly except maybe the aggressive tone ^^

[devius]

How can Gnome be intuitive? KDE uses the standard desktop metaphor that's been in use for almost 20 years. There's nothing more intuitive than that.

It's not the standard desktop metaphor, it's the Microsoft Windows95 desktop metaphor. That's like saying right-hand traffic is the standard in driving. Or that the Nintendo DS is the standard mobile console.

What I mean is that it may feel more easy to get used to when coming from a Windows-centric world, but it's not the only desktop paradigm that exists. Anyway, that "standard" will die a little with Win8 so the world is clearly moving forward. Nothing ever stays the same forever.

[liam]

The problem is, the offered paradigms don't "scale" if you have many windows open.

Sure, if you just have Firefox and Rhythmbox open, alt-tabbing is fine.

But if you have 10 or 12 windows open, the absolute easiest way is to commit to muscle memory exactly where the window is relative to the position of the others in your taskbar, and just move the mouse down there and click it. It takes about 500 msec instead of "Alt-Tab... nope, not that one... alt-tab... nope, not that one... alt-tab... nope, not that one... alt-tab... nope, not that one... THERE we go!". And if you want to use the mouse, you have to move it all the way up to the top of the screen, then wait while it renders an iconified version of all the windows, then move your mouse to click on the one you want. That takes a lot of time. The other problem with the window selection is that the tiling feature "wraps" the windows around to the next line, so it's easy to lose track of the position of one window in the list. So for frequent task switchers (*raises hand*) it's useless.

The problem is the taskbar doesn't scale either, even assuming you are grouping by application. Even if it did, it isnt clear to me that "muscle memory" will scale.
Alt-tabbing works well, and is very fast, b/c it lets you use more of the screen, and overview scales even better since it uses most of the screen, but as you add more apps each app gets smaller and let recognizable.
The only thing I know that will scale is group by task. Virtual desktops are a primitive way to achieve this but require too much manual work to both setup and use. KDE has Activities, which is a more refined version that lets you name and customize each space but is not very well designed (from a ux standpoint, but this and bugs have always been the problems with KDE). Something similar to that, but offered in a more user friendly way, would be fantastic. The obvious GS method to do this would be to get rid of the virtual desktops, allow dnd of app windows to create groups, and name each group. This is what firefox does and it scales better than anything I've ever seen and is very intuitive to use, but it isn't very discoverable. Adding that to Overview fixes the discoverability.

[fuzz]

The problem is the taskbar doesn't scale either, even assuming you are grouping by application. Even if it did, it isnt clear to me that "muscle memory" will scale.
Alt-tabbing works well, and is very fast, b/c it lets you use more of the screen, and overview scales even better since it uses most of the screen, but as you add more apps each app gets smaller and let recognizable.
The only thing I know that will scale is group by task. Virtual desktops are a primitive way to achieve this but require too much manual work to both setup and use. KDE has Activities, which is a more refined version that lets you name and customize each space but is not very well designed (from a ux standpoint, but this and bugs have always been the problems with KDE). Something similar to that, but offered in a more user friendly way, would be fantastic. The obvious GS method to do this would be to get rid of the virtual desktops, allow dnd of app windows to create groups, and name each group. This is what firefox does and it scales better than anything I've ever seen and is very intuitive to use, but it isn't very discoverable. Adding that to Overview fixes the discoverability.

GS already has app groups... but they are currently based on programs. Of course virtual desktops act sort of like groups of groups, and there are keyboard shortcuts for all of them

Ctrl + Shift + arrow keys to switch virtual desktop (and diagonal works too)
Alt + Tab to switch program groups
Alt + ` to switch windows between programs

It's amazingly efficient.

People would know if they just read the really short handbook.

EDIT:
And these wont all fit on a task bar .

[liam]

GS already has app groups... but they are currently based on programs. Of course virtual desktops act sort of like groups of groups, and there are keyboard shortcuts for all of them

Ctrl + Shift + arrow keys to switch virtual desktop (and diagonal works too)
Alt + Tab to switch program groups
Alt + ` to switch windows between programs

It's amazingly efficient.

People would know if they just read the really short handbook.

EDIT:
And these wont all fit on a task bar .

If you read my post you'd see why I think virtual desktops, as implemented, don't work well. Also, I don't understand what you mean by saying that GS currently has app groups but are program based.

I am very familiar with the short cuts ever since Marina started documenting them well before the first GS release. I like the shortcuts but the current arrangment doesn't make it easy to setup task based work flows. It can be done but I have to force myself to do it. The changes I spoke of would make things much simpler. Then you could alt tab between tasks and alt ' between windows, or whatever.
People shouldnt habe to read a handbook in order to use a desktop efficiently. The shortcut overlay that is planned will help alot in this area, but most users should only need to know a couple in order to work quickly. Vim/emacs style is great once you learn it, but it takes to do so. A desktop user shouldn't need such dedication in order to use a desktop.

[Danny3]

You are in luck, I happen to know of such a distribution. I call it Fuduntu (http://www.fuduntu.org).

Thank you. It's based on Fedora so it uses rpm, but it doesn't matter, I see it like an opportunity for me to learn about yum and rpm.

[johnc]

It's not the standard desktop metaphor, it's the Microsoft Windows95 desktop metaphor. That's like saying right-hand traffic is the standard in driving. Or that the Nintendo DS is the standard mobile console.

What I mean is that it may feel more easy to get used to when coming from a Windows-centric world, but it's not the only desktop paradigm that exists. Anyway, that "standard" will die a little with Win8 so the world is clearly moving forward. Nothing ever stays the same forever.

I don't think Windows 8 is a good example of "moving forward", at least from a desktop / laptop perspective.

It's okay to come up with a new paradigm, but it has to be better than what currently exists. Simply being "new" and "different" certainly doesn't cut it.

[randomizer]

It's okay to come up with a new paradigm, but it has to be better than what currently exists.

"Better" is subjective. What metrics do you use to determine what is better (or in what way it is better)? Oh right, that's subjective too.

Simply being "new" and "different" certainly doesn't cut it.

No, but they're prerequisite to being better.

[johnc]

"Better" is subjective. What metrics do you use to determine what is better (or in what way it is better)? Oh right, that's subjective too.

It's definitely subjective. Which is why the user base will sort it out eventually.