GNOME: Possible Recovery Strategies

An anonymous reader tips an article from Datamation about several suggestions for the GNOME project to answer user complaints and boost developer morale. From the article: "... with very few changes, GNOME 3 could be much more acceptable to most users. A moveable panel, panel applets, desktop launchers, user control of virtual desktops, menu alternatives that would remove the need for the overview -- all of these could be added easily as options. Together, they would reduce at least ninety percent of the complaints against GNOME 3. ... If GNOME is having trouble as a desktop environment, one obvious solution is to find new niches. Lopez and Sanchez suggested following KDE's lead and producing a tablet, while Lionel Dricot recently suggested a suite of cloud-based services. ... The one strategy that GNOME has never tried is asking users what they want. Instead, the project has preferred to rely on usability theory, treating it as an exact science instead of a collection of competing ideas supported by usually inconclusive studies that could be mustered to support almost any design. In GNOME 3, testing with actual users did not occur until near the end of the development cycle, when the chances of any major changes were remote."

Comments

Staying with gnome2

[DCFusor]

Because 3 sucks and they don't listen to real users. Theory ain't the same as practice, in practice.

Re:Staying with gnome2

[gagol]

OpenSource economics are different from commercial ones. The end goal is offering something unique that will appeal to a set of users. The fact that you can install a DE from many providers means people who prefer traditional desktops can turn to LXDE/XFCE, if you want eye-candy and "paradigm" buzzwords, you can use KDE/GNOME, you prefer a tiling desktop, install AwesomeWM, etc...

My point is, the end goal is to fill the niche, GNOME3 try to fill them all and failed to find it's sweet spot...

Re:Staying with gnome2

[buchner.johannes]

Because 3 sucks and they don't listen to real users. Theory ain't the same as practice, in practice.

The largest screwup was not by GNOME but by distributions in my opinion.

They abandoned GNOME3 for GNOME2 after it was released, not bothering to offer both choices. Some like Gentoo do provide the choice, for bleeding-edge distros like Fedora I understand that they went with the newest. But user-distros shouldn't have gone for GNOME3 only, and there is no technical reason to not offer both.

I think GNOME wants to build a interface for users and not for developers, which is why the slashdot community is a bit pissed (not being the target audience, complaints about "dumbing down".

KDE is elaborate and clunky; XFCE is a good tradeoff; more minimal WM are just toys for having multiple terminals. The choices offered to users by distributions was better a few years ago.

Re:Staying with gnome2

[Derek Pomery]

According to the MATE developers (the guys maintaining a fork of Gnome2) there is one technical reason you can't have both.
Gnome is not designed for multiple versions to be installed simultaneously. There are name collisions.

You can have multiple versions of KDE installed. Not Gnome.

According to the MATE devs ( irc://irc.freenode.net/mate ), that's why they had to rename pretty much everything.

I can attest that MATE and Gnome3 *can* be run on the same machine, although MATE is still getting on its feet.

And some distros are offering both.

Re:Staying with gnome2

[Electricity Likes Me]

I blame the MS Ribbon. It was the first shot fired in the modern era of UI redesigns, wherein it was decided that the real problem was there was no way to force user's to use it until they like it.

Re:Staying with gnome2

[epyT-R]

You don't make a case for the gnome 3 changes here. You just make assumptions about the people who criticize it. Old stuff isn't necessarily worse than new stuff, and new stuff isn't necessarily worse than old stuff. They both must stand on their merits. This trend of minimalism in modern UIs and applications was fine until they started cutting needed features and/or flexibility for its sake. Gnome 3 is doing this along with windows 8, and osx. I'm sorry, but I don't need all these assumptions made about where I keep my windows on a workstation class machine. They are not tablets.

Change for the sake of change isn't innovation.

Re:Staying with gnome2

[techno-vampire]

Or you can try actually learning the new system - it really is better.

What you mean is, "It's better for me." I want to be able to put my panel where I want it, not where the devs want it and I don't think I should have to install a third-party extension to do that. I don't want to have to use gestures to get to a list of applications, I want to use both icons and menus. I want to be the one who decides how many workspaces I have, and what programs appear on which. AFAIK, none of those things are possible in Gnome 3, which is why I now use Xfce, where they are. I might add that after a year of fighting with Ubuntu's Unity DE, which is pretty much a clone of Gnome 3, my sister gave up on it and now uses Xfce as well.

Re:Staying with gnome2

[RocketRabbit]

Gnome3 is garbage unless you have a tablet. It's a tablet UI, designed without considering that most people who run Gnome do not use a tablet. It's not exactly elegant even compared to other tablet UIs, but it's better than Gnome2 in that respect. However, as I said, people don't really use Gnome (ANY Gnome) on their tablets.

If you're happy with it, great. It's clear that most people aren't, and no matter how much you insist they are stupid for disliking something THAT IS COMPLETELY SUBJECTIVE they won't use it.

Re:Staying with gnome2

[tolan-b]

I'm using it as my DE and have been for a few months. While there are some nice features there are a lot of serious problems and the complete lack of customisation outside of extensions (which break on each point release and need to be updated manually, if theres a version compatible with the new release) makes it very frustrating. Having the activities button on the top left is terrible for me because i dwell the cursor on the right (window controls, scrollbars are on the right). The new task-tray equivalent is frustrating because it takes longer to get at things like music player controls, and the way the items expand makes them hard to click. Needing to know you have to hold alt to access the shutdown function (instead of sleep which I've still only had work on one computer in Linux) is terrible from the pov of discoverability. The use of a fixed hot corner is terrible for multi-monitor, and I still dont understand how the workspaces work on a second monitor, its pretty bizarre. Also while not necessarily a bad design, I personally hate the launcher and would prefer a menu. Also the lack of applets is annoying, having disk and cpu usage is very useful for me.

Much of this would be solvable if some configurability were built into the shell but the overwhelming "we know best" philosophy now prevalent in gnome makes this impossible.

While I do like a fair bit of the UI in gnome I'm sick of fighting it and as soone as Mate is a bit more mature I'll be moving to it and having a (configurable) hot corner for 'expose' in compiz... I do hope Mate moves to GTK3 though, as I like the toolkit.

Re:Staying with gnome2

[MysteriousPreacher]

until they started cutting needed features and/or flexibility for its sake. Gnome 3 is doing this along with windows 8, and osx.

Which features/flexibility have been cut from OS X for the sake of minimalism?

Re:Staying with gnome2

[gbjbaanb]

I also blame MS, but not for the ribbon (that's just a symptom of a greater underlying disease). The reason MS has dumped the start menu - because their usability labs have decided that users don;t actually use it, 90% of users preferring to stick icons on the desktop (you've seen them) or pinned to the taskbar.

Now while that is undoubtedly true, and shows that quick-access to often-used programs is a very important feature, it forgets to note that people still use the start menu for all apps that are not quick-launched. But, hey, that doesn't matter, the last 10% of user activity can be sacrificed in the name of statistical user input.

Same with the ribbon - its basically a quick-launch menu, only forgetting about the bits you do not use often.

It seems Gnome has the same problem, focussing on a flawed assumption that if a user doesn't use something all the time, then they don't use it at all.

Re:Staying with gnome2

[Grishnakh]

Just because lots of time was put into engineering it does not make it good.

Exactly. A lot of time was put into engineering Windows Vista, as well as Windows Me, and those sucked. A lot of time was put into engineering the Ford Pinto, and that was a deathtrap. A lot of time was put into engineering the Pontiac Aztek, and that was so butt-ugly it's known to many as the ugliest car ever made in all history. A lot of time was put into engineering lots of "enterprise" software, and that entire class of software is frequently scorned as "crap" among technophiles here.

Re:Staying with gnome2

[RocketRabbit]

I'm sorry but you'd have to be high on crack to believe that anything but a teeny tiny minority of Linux users like Gnome3. People hate it. Never before has there been a desktop environment that has received such universal dislike.

"Thank you Gnome3 team" indeed. You probably are one of the developers that foisted this abortion upon the world. In fact, I am beginning to suspect that Gnome3 is a Microsoft funded project to scare people away from Linux.

Re:Staying with gnome2

[RocketRabbit]

"I seriously just don't get the hostility. Is this a case of the windows refugees getting restless?"

No, this is a case of legions of loyal Linux users getting pissed off because their UI has been fucked with an iron stick with a razor edge.

You don't get it - the Gnome3 team pushed their vision and now the users are pushing back. Never before has there been a fork of an old version of a DE that received so much attention. Mate is Gnome, as far as anybody with a clue is concerned.

You talk about keyboard-centricity, yet I'm sure you noticed that Gnome2 had the same functionality. Gnome3 has removed a whole lot of choice that the user had, and sure - you can fuck around and download a whole bunch of extensions in order to make it more like Gnome2. In fact, there are countless guides online devoted to getting the Gnome3 disease under control.

The masses have spoken. They hate Gnome3.

Not just Gnome

[Meshach]

The one strategy that GNOME has never tried is asking users what

Almost all software has that problem.

Re:Not just Gnome

[smpoole7]

> Almost all software has that problem.

This. Especially among open source projects. I deeply appreciate their efforts, but when you go into their forums with a suggestion, or to ask why they are doing something a certain way (or more often nowadays, why they stopped doing something that everyone liked), you get scolded. Or talked down to. "Trust us, little man, we're the experts and we know what we're doing."

This article is about Gnome, but I'm still sore from the way the KDE developers handled their transition to version 4. Even the politest request was greeted with outright hostility. Gnome is by no means the only offender, nor is the offense limited to desktop environments. But it's a real problem.

I much prefer open source to proprietary software, but there's a price for the "free" stuff. I guess this is just part of it. A commercial software product that treated its "customers" the way that some FOSS projects do would be out of business in a matter of weeks.

Just my opinion and worth exactly what you paid for it. :)

Re:Not just Gnome

[Anonymous Coward]

In that analogy, trying to put a tablet UI on a PC desktop environment would be like trying to put a steering wheel on a horse.

Re:Not just Gnome

[Anonymous Coward]

Yeah, Apple and Microsoft have never pissed of thousands of users by redesigning GUIs and ignoring complaints.

Re:Not just Gnome

[ShieldW0lf]

Writing software is not "art". It's not there to be appealing. Writing software is about building tools, and when you're dealing with tools, the "right way" exists.

I've written a ton of commercial software, and if you're going to do it right, the first step is convincing the customer that they don't know what they need, and that your very first task will be interviewing them so you can give them a document that tells them what they need.

If you can't convince the customer of this truth, you're usually better off firing the customer. Not only do they usually take more time to deal with than the money justifies, but they actually train you to become shortsighted and less effective and leave you slightly crippled moving forward.

None of which means I like Gnome 3... but this stupid desire to abandon the WIMP metaphor and transform the powerful tools we know and love into crippled and useless entertainment devices seems to be industry wide, and neither a Gnome specific problem nor an open source problem.

The screwball thing about it all is, Windows 8 is coming down the pipe, it looks to be just as fucked up as Gnome 3 is... and the open source community no longer has a decent and well supported WIMP desktop available to capitalize on the opportunity.

Re:Not just Gnome

[jbolden]

Sure it does. KDE (which is really quite good). Cinnamon (a fork of Gnome 3), Mate (Gnome 2) and possibly XFCE or LXDE.

Re:Not just Gnome

[gagol]

Try XFCE4... you will be surprised.

Re:Not just Gnome

[UnknownSoldier]

> Writing software is not "art".
Sorry, you're wrong. Yes, there is a lot of science in Computer Science, but since this topic is about UI -- as soon as you start interacting with users, there are times when it is OK to break the UI rules. The *hard* part is knowing when to be consistent, and when not to. People, nor how they interact with computers does NOT always fit in a nice little black-n-white box that naive programmers love to think.

And just to be pedantic, here is real-world example: (Since /. is a POS for code formatting, replace the _ with spaces...)

The most important thing for writing code is: proper variable names, whitespace to align common idioms

function SwapInt32( x )
{
        var n _= (x >> 24) & _____ 0xFF;
            n |= (x >>_ 8) & ___ 0xFF00;
            n |= (x <<_ 8) & __0xFF0000;
            n |= (x << 24) & 0xFF000000;
        return n;
}

Proper alignment makes it easier to read code. There are no hard and fast rules for whitespace.

> It's not there to be appealing.
Methinks
a) you missed the joy of optimizing code and coming up with a smaller and faster algorithm, nor
b) even grok the purpose of whitespace in the first place. Hint: Whitespace is NOT for the compiler's / interpreter's benefit but _humans_.

> the first step is convincing the customer that they don't know what they need,
Yes we understand your point that "No, the customer is not always right".

But riiiiight, like the customer is always some clueless schmoe. News flash, sometimes, they have been using software *longer* then your little code monkey shop has been in business for. While they may not know exactly what they want, it pays attention to try to understand their perspective and what are they *really* getting at. One of the best ways to learn how bad your UI is, is to give it to someone who does not have the same preconceived ideas that you automatically *assume* all your clients and other programmers have.

In the *real* world, *sometimes* client ARE knowledgable -- AND sometimes they are completely clueless. Your job as a programmer is to bridge that gap, and learn to get at what they are *really* wanting.

If you think programming is black-n-white you obviously haven't been doing it very long, or you suck at it.

Re:Not just Gnome

[MikeBabcock]

No offense, but I hope I never have to use your software.

User interfaces are all about art. A right way doesn't necessarily exist. Is right clicking better than a button? Are four buttons too many, or is seven? How many view types should be on one screen?

These vary from system to system, function to function, and a piece of software may work perfectly but suck because the user can't use it efficiently or simply hates using the software.

Lots of picky examples exist from the mundane like when I mouse over the chat window in Facebook, I expect the chat window to scroll, not the main window, when I roll the mouse wheel -- to the customer I have who want Enter to go to the next field in a form not tab because that's how it would work on a spreadsheet or a calculator.

Form shouldn't override function -- but form is very important, and almost entirely art.

Re:Not just Gnome

[spauldo]

Microsoft's done it precisely twice. Ribbon, Metro.

Add the Windows 95 / NT 4.0 UI to the list as well. I know a lot of people here diss Windows 3.x, but at the time a lot of users hated 95 and wanted to go back.

(I was one of them, not because of the UI, but because 95 was a buggy piece of shit)

Oh, and Bob would have counted if they could have gotten the computer companies on their side. A few actually included it as the default interface.

While it's a lot more minor, I personally also include the blue candy look that XP defaulted to. I hated the default XP theme and always changed it to the classic theme on any machine I had to work on. Fortunately, by the time XP was widespread, I no longer worked with Windows :)

Meanwhile, open source?

The fuck is wrong with you people?

Lack of strong leadership, generally, although you also have to account for trends in UI design that evolve over time. Today, most people run GNOME or KDE. Back in the day, it wasn't like that at all - commercial users used CDE (unavailable for free software) or OpenLook (which never gained much traction on free software for some reason). The free UNIXes had a bunch of different window managers along with a bunch of different toolkits. Most of us thought OpenSTEP would take us away from Athena and Tk, but that never really happened. Window managers explored a bunch of different ideas, and Enlightenment was going to make the world a better place if only anyone could afford a machine that could run it.

KDE and GNOME came along and unified a bunch of stuff, but in doing so you lose choice and control over your desktop. GNOME especially has tried to remove options and configurability to try to appeal to some hypothetical end user who couldn't be trusted with sharp objects and tended to try to eat rocks. GNOME 3 is a logical extension of this philosophy.

Here's the kicker: you don't have to play. You can install your distro's GNOME and deal with it, or you can find something you like that's stable and stick with it. I use a highly customized FVWM setup that hasn't changed significantly since about 1998.

Re:Not just Gnome

[fnj]

The one strategy that GNOME has never tried is asking users what

Almost all software has that problem.

Those are both very perceptive remarks by the article and the poster. The difference with Gnome is that they not only (1) didn't ask, but also (2) went with stupidly awful ideas of their own. Obviously you can survive (1) but not if (2) is also active. And not all software projects have stupidly awful ideas of their own. A lot of them do, but nowhere near all. For example, Xfce is dead on the mark with their instinct.

Re:Not just Gnome

[fnj]

Sorry, Bob and the Paperclip DO count. It's absurd not to count them (even though I personally never hated the paperclip with any fervor, the general verdict in in). And the pervasive awfulness of Windows Me. And the hopeless morass of Windows Vista. And the garbage heap that is Windows 8. You can't just excuse away their continual stream of massive goofs.

On the other hand, open source works, you anonymous idiot. So Gnome lays an egg; so what? It doesn't automatically make linux stink, because there is a wealth of lines of development going on, not like the walled garden of closed source. Just switch to Xfce for god's sake. If Xfce and half a dozen other alternate solutions were not already there, SOMEBODY WOULD START THEM. Because they COULD.

Re:Not just Gnome

[smpoole7]

> One of the best ways to learn how bad your UI is, is to give it to someone who does not have the same preconceived ideas that you automatically *assume* all your clients and other programmers have.

Another thing I have suggested for ANYONE writing software -- whether FOSS or proprietary -- is to hand your software package to an end user. Then go sit in another room and watch them through a window. You can't help them or give them tips. Watch whether they struggle with it.

I've done this myself and 15 minutes watching a real, live, end user is more profitable than anything I can think of. Speaking from experience, the first thing you'll likely discover is that there are libraries on your development machine that aren't on the end user's, and you forgot to include them in the package (even if only as listed dependencies for the package manager). But once you get it installed, you sit back and watch. Look at their frustration as they try to figure out which menu items to click to do what they want.

More often than not (I've seen this, too, unfortunately) is they'll just give up and go back to what they're used to. If they can't easily navigate around your Brand New Thing(tm), they're going to blow it off.

I think that if everyone who developed software would do this simple bit of research, it would be a much happier world. :)

Re:Not just Gnome

[epyT-R]

The problem with today's software is that these aesthetics are taking more precedence than functionality and usability. for example, gnome 3's window management makes egregious assumptions about what I want my windows to do. Just because I move a window to a certain spot doesn't mean I want it radically resized. Something reasonable would be snap-to-grid or to-edge, which is nice, WITH an option to disable it if it causes a problem.

There is a reason for such hatred, and that reason is likely embedded in the workflow assumed by the software. Today's modern UIs are rife with this sort of thing..the looks matter more than the workflow, the latter being designed for mouth breathing idiots. I realize this is a necessity for input limited devices like tablets, but it does not belong on workstations. These 'designers' know this, but they'd rather cash in on stupid fads and hot trends than develop good software, or in the case of gnome devs, brownnose apple.

The problem with facebook and other web 2.0 'applications' is that the browser was never designed to handle the sort of contexts you're referring to. Money and control freakery drive 'web apps,' not good design, aesthetics, or user interest. Use proper tools for the job, in this case, a real IM client.

Re:Not just Gnome

[Sipper]

Have you upgraded to 4.8 or 4.9, which I heard is a lot better? Or do they still have similar problems w/ Nepomuk and Strigi?

I'm running KDE 4.8.4, which is what is in Debian Unstable. Before a presentation I did on KDE4 for my local lug I tried Strigi/Nepomuk features again in KDE4.8 and performance was again terrible -- many hours of 100% CPU time during the indexing process. [IIRC on the same P4 system this process took somewhere between 14 to 18 hours to index a home directory with 30 GB of stuff in it, and I think the resulting Virtuoso database was about 1 GB.] The reason is that Nepomuk/Strigi uses several "ontologies" run as separate background processes to do the indexing -- one "ontology" for each type of indexing being done -- file names, pictures, audio files, etc -- so you'd think this would be a single background process searching the disk, but no it's like 5 or 6. And the thing is, I have no reason to use either of these services. To even find out what these services can do isn't easy, because the documentation for them in the KDE manual is terrible. However if you search around the 'net you'll eventually find this:

https://kdenepomukmanual.wordpress.com/2012/02/06/detailed-kde-nepomuk-manual/

As you read the above document here's some details to keep in mind: I use Krusader as my chosen file manager, not Dolphin. I don't use DigiKam or Gwenview (I use Geeqie as my chosen picture viewer). I never use the Alt-F2 Krunner menu, and I never give files tags or comments to them. Therefore Nepomuk as it stands today is totally doesn't serve a purpose for me.

But above all else, I don't want KDE4 choosing on its own when to run a super I/O intensive indexing process just to create a *cache* of things it finds and hold all of that in a huge database. To me that harks back to the 'mlocate' package -- if you've ever been working on a server that suddenly ran like shit and you found an 'updatedb' process running when you viewed the output of 'top', that's the package that did that. But unlike the mlocate package that called the updatedb process via cron, there's no way to tell KDE4 when to allow it to do the indexing or to limit CPU or I/O -- the only thing you can limit is how much RAM is used, which doesn't address this problem. The indexing starts by default, immediately after your very first login, causing the computer to run like utter shit, and your only choice to stop it is to immediately go to System Settings -> Workspace Appearance and Behavior -> Desktop Search and turn these features off . And that's only if you know where to go and what to do. This is why many users that try KDE4 for the first time say "I can't use this, it makes my computer run like shit." And unfortunately, by the default settings KDE4 gives you, they're right.

There are many other levels of FAIL here, too -- the listing of Control Center modules in the KDE Help are not even in alphabetical order, so even if you somehow know that settings for Nepomuk are in the "Desktop Search" section, it's still an effort to find in the list. Then once you look through the help for the Desktop Search, the documentation that is there is simplistic and doesn't even tell you how you can use it, and doesn't give you any warning whatsoever of the performance impact these services have. [I discussed the lack of performance warning with the KDE4 developers at the time, and they were again unsympathetic. After about a week of arguing and "talk to the hand", they told me to create a KDE account and to propose wording myself, which they'd then review and consider. The problem with this suggestion is that they had already made it quite clear that they were not going to take it seriously.] And the way you get into trying to fix the performance disaster is finding several 'nepomuk' processes, so you try to fi

Re:Not just Gnome

[vurian]

"they basically just dismissed 3.x, said "we ain't doing it no more". That is just not true. We released two more releases of KDE 3.5 after KDE 4.0 was released.

Re:Not just Gnome

[UnknownSoldier]

Sure.

The classic paper every programmer (IMHO) should read is one by Danny Cohen who introduced the terms big-endian and little-endian is:
    ON HOLY WARS AND A PLEA FOR PEACE
http://www.networksorcery.com/enp/ien/ien137.txt

Here is a "abridged" summary:

Like most things, there are 2 (diametrically opposing) way of doing things. You can do it either A xor B; that is A, or B but not both. Neither is right; they are just (competing) standards. Oblg. http://xkcd.com/927/

Agin, neither is technically "right" - they both are; we may chose one or the other *simply* for convenience. As long as we pick one standard, and are consistent in our use, everything works. The problem arises when somebody picks a different "standard" and we need to interface with them. (i.e. share data.) :-)

Here is practical example that shows up all the time in computer graphics. Let's say we wish to define the world coordinate system. To the right we can call that positive X, we can call up positive Y, and for stuff off in the distance we have a choice:
a) call positive Z for things further away from us, (DirectX uses this) OR
b) call negative Zfor things further away from us (OpenGL uses this.)

Getting back to the function originally mentioned:

When we have the integer number 0x12345678 in a CPU register how the bytes are stored in memory / disk / "network" can be done either in:
a) Little Endian format:
i.e.
    unsigned char LE[] = { 0x78, 0x56, 0x34, 0x12 };
or
b) Big Endian format:
i.e.
  unsigned char BE[] = { 0x12, 0x34, 0x56, 0x78 };

How did this come about? Notice that we write bits from right-to-left
    bit31 bit30 ... bit 3 bit2 bit1 bit0
the same as we write the Aramaic numbers. Now which bit should we send first over a wire? bit31 or bit0? The hardware guys wanted it one way (because they could use a simple barrel shifter/latch), the software guys wanted it the opposite way.

Let's say you have this number stored on disk that was generated by a program running on a little endian CPU. If you have a big endian CPU try to naively read this data it will interpret it as the wrong value.
i.e.
  Little endian 0x12345678 (305,419,896) on disk is: 0x78563412
  Big endian will interpret it as: 0x78563412 (2,018,915,346)

Hence you need to "fix up" the bytes, that is byte swap them. The CPU instructions of shift-and-mask is the standard way to shuffle bits around.

It helps to label all 32 bits with unique names:
    ABCDEFGH IJKLMNOP QRSTUVWY abcdefgh

We want a function that, given the above, will generate this:
  abcdefgh QRSTUVWY IJKLMNOP ABCDEFGH

Or expressed in bytes:
    b3 b2 b1 b0
We want:
  b0 b1 b2 b3

Notice that the hex mask 0xFF (called a "byte mask") is a way to treat 8 consecutive bits (one byte) as one "logical number".

By inspection:
  b0 should be shifted Left 24 bits
  b1 should be shifted Left 8 bits
  b2 should be shifted Right 8 bits
  b3 should be shifted Right 24 bits

When you are talking about numbers in base 16, the idiom << 8 of shifting the bits by one byte, or multiplying by 256, is equivalent to a multiplying a decimal number by 10 because we want to shift all the digits over to make room for another "tens" unit place.
i.e.
  0x12 8 = 12 times 2^8 = 12 times 256 = 0x1200
  12 * 10 = 120

Does that answer your question sufficiently?

Re:Not just Gnome

[spauldo]

I completely agree with the whole argument on mouse/keyboard thing. I remember when light pens were around (a miserable failure for general use) and we've all used touchscreen systems that were a pain in the ass compared to a mouse/keyboard setup. I don't have a problem with an interface that is mouse-heavy (I don't use keybindings in FVWM, for instance, relying solely on the mouse for window management), but it seems silly to me to try to make a computer interface resemble a touchscreen interface - they're completely different in the way you interact with them.

Now the off-topic part: My FVWM config isn't all that to look at. I use the old style MWM look, which I've always found appealing (but then again, I think QT is ugly and I like the look of Motif, and I've come to terms with having bad taste). I've used the same desktop background for over ten years.

My customizations have all been mouse and window behavior stuff for the most part. All the buttons (the default MWM-style one on the left and two on the right) do different things depending on what mouse button you click. For instance, rather than a close button on the right, you right-click the menu button on the left (I've always tended to the upper left for close since Windows 3.x, and it's annoying that some windows don't have it in newer Windows versions). You get window management functions by left clicking (move, resize, close, etc.). A right click on the title bar shades or unshades. On the right, the iconify button does pretty much just that, going to an iconbox above my FVWMButtons panel (set as 64x64 boxes running vertically on the lower right to get a bit of NeXT look). I rarely iconify anything, though, since shading is so much easier. The maximize button maximizes vertically for left click, with a full maximize on right click. I've set the maximum window size so that my FVWMButtons panel doesn't get covered on maximize, but it's not set to be always on top, so I can cover it if I want to. There's no autoraise - I like being able to type in one window that's underneath another window.

You can move the window by middle-clicking any border or just left clicking the titlebar. Resize is left-click on a border. Right clicking the border will either bring the window to the top or put it on the bottom. Moving the window never changes a window's position in the stack.

I use edge scrolling, which is a feature that's all but disappeared in other window managers. I also use a 3x3 workspace grid (workspaces and desktops behave differently in FVWM, and I find workspaces work better for me). There's a pager in the FVWMButtons panel. I've got a button that toggles scrolling on and off for when my 4 year old wants to play with google maps or watch trains on youtube.

I use xterm. For years I played with rxvt and eterm, playing with transparent backgrounds or pixmap backgrounds, but I don't really care about that these days. Xterm has unicode support and is easy to configure on the fly while still being loads faster than the "smart" terminals that come with GNOME and KDE. I don't need multiple sessions in one terminal (if I did, I'd use screen) and xterm has tons of configuration options in the .xresources file.

Most of my apps are GNOME or GTK+ apps, and for the most part they work just fine. They just use the default theme, since I don't really care what they look like. I still use a few old Athena programs (xcalc, xfontsel, xclock, etc.) and lament the fact they're going away (I like X resources. You can do amazing stuff with them. I had Netscape 4 (Motif-based) with pixmap backgrounds on the UI years before anyone tried theming Mozilla). I'd really like a nice Athena-based battery monitor I could swallow into FVWMButtons, but Linux keeps changing the interface for it and none of the old ones work anymore.

OK, end of off-topic crap.

What Gnome 3 Needs

[rcjhawk]

Is a big button on the panel that says "Make it Work Like Gnome 2" Or FVWM, I'm not picky.

Re:What Gnome 3 Needs

[2.7182]

Dead on. FVWM especially. I regret the day I left FVWM, thinking there was something better out there...

Re:What Gnome 3 Needs

[Ruie]

Is a big button on the panel that says "DO NOT PANIC" and makes it work like Gnome 2.

FTFY

Re:What Gnome 3 Needs

[epyT-R]

Unity keeps things simple the same way an etchasketch keeps image editing simple...

Extensions

[macemoneta]

The requested functions are already mostly available via gnome shell extensions, allowing users to customize gnome to their preference.

Re:Extensions

[redmid17]

Great point. Everyone prefers a piece of shit out of box that you have to shine and polish to make look nice.

Re:Extensions

[Anonymous Coward]

The functionality is available as Linux comes with a C/C++ compiler.

Re:Extensions

[93 Escort Wagon]

That could work for most /. users but most regular users neither know how to enable extensions or care enough to learn.

We're talking about desktop Linux here - "regular users" aren't really a concern.

Re:Extensions

[Anonymous Coward]

The requested functions are already mostly available via gnome shell extensions, allowing users to customize gnome to their preference.

And this is where they fail. No one wants to program a fucking extension for every little bit of "useful" feature that should be there right out of the box so to speak. And that by virtue of being an extension could go away anytime. It's the same disease that affects the Firefox developers. Until this simple concept is hammered inside the gnome-tards thick skulls the project will remain a big fail.

Re:Extensions

[sqldr]

I'm one of those rare people who stuck with that piece of shit and actually got the hang of using it efficiently. None of the suggested windows-95 throwbacks in the article are things I WANT back. I install about 5 extensions out of the box, and the only "tweak" I use is turning on focus-follows-mouse and making better keyboard shortcuts for desktop switching. The auto desktop management thing is a really efficient way of working once you get used to it, rather than assigning 4 desktops to different activities, then after an hour of use realising you've been putting the wrong windows on the wrong desktops and that you've got shit everywhere. It makes you think where you put your windows.

Re:Extensions

[MikeBabcock]

I don't want to think where I put my windows. I know my personal browser sessions are on 3, along with any game I might be playing, my E-mail and other contact managers are on 1, and my database interface and Eclipse are running on 2.

When I want to save a window for later, I toss it over to 4.

I shouldn't have to think about it. That's how proper organization works.

Imagine for a moment if your clothing drawers automatically created and deleted drawers so you had to figure out where you'd put something, and if you took the last sock out of the sock drawer, the shirt drawer wouldn't be where you expected it. We use metaphors on desktops to help users organize their data, including the folder system. Making those metaphors less realistic kills their ability to use them for organization.

Re:Extensions

[damnbunni]

Most people I hang out with keep their phone in their back pocket.

'Year of the Linux Asstop' doesn't have quite the same ring to it, though.

Re:Extensions

[cfalcon]

What the fuck is a static workspaces extension?

Fucking shit the last time I had to grab tinyturdware in order to have a useable GUI, it was before Win95. Now I need a list of downloadable crap that ONLY EXISTS BECAUSE THE BASELINE SUCKS ASS, and I don't even know the list. Maybe there's something to browse?

Whatever, fuck it, and if you defend it, fuck you too. The sheer attitude of the devs is so hard to explain. It's like, they build it for some lowest common denominator that doesn't even exist, and then if you don't like that you must be some kind of problem case so go get a dumb extension? I guess it's good that they have those now, when I left that GUI they sure as shit didn't, they just had a bunch of goddamned attitude.

Fuck it. Just fuck it. GNOME is a lost fucking cause until it gets forked by devs that don't have their heads so far up their asses that they are topologically equivalent to a fucking klein bottle.

They won't listen anyway

[Anonymous Coward]

GNOME devs are not going to aknowledge their mistake. No, for them, it's everyone else who are mistaked about the way they should handle their work. And, of course, it's GNOME devs who know it best. Their design is marvelous, all that is left is for user to bend himself to it.

That's why GNOME 3 is stripped of so much functionality, deemed "unneeded" by devs on the basis of them not needing it. And they continue upon this path: http://blogs.gnome.org/mccann/2012/08/01/cross-cut/

KDE has it, too, but to a lesser degree and most of the time they let user configure his environment.

Re:They won't listen anyway

[Junta]

Holy crap:

A lot of reasons people have been using this view are due to the other two views sucking for various reasons ... The role for compact view is unclear. Our research suggests that it is something like: the only view that works for browsing a lot of files at once. This is really hard to reconcile with providing good defaults that just work and having consistency with the file chooser.

So you admit people are using the view, it works best for browsing lots of files, and somehow, this means the reason for existence is unclear somehow so you should delete it because you don't use it yourself?

Meanwhile, they try to circle the wagons and discuss what to do to address an issue of dwindling support:
http://www.phoronix.com/scan.php?page=news_item&px=MTE0ODg
Their conclusion including how to address brain drain and exodus of users? *MORE* Gnome 3, stop thinking about the desktop paradigm as much and make it more different, and Gnome hasn't taken over *enough* and needs to be its own OS.

Oh well, guess GNOME will descend into oblivion. They had some neat aspects in Gnome 3, but it's just so hard to deal with some of the intended design choces that they clearly have no intention of revisiting.

Revert back to what worked

[Anonymous Coward]

GNOME 2 wasn't broken when ivory tower developers decided to fix it.

Why not spend development resources optimizing accelerated graphics performance and squashing bugs?

Don't screw up the perfectly fine UI because you have nothing else to do. (GNOME 3)

Don't bloat the whole DE beyond belief and require users run multiple heavy daemons with a questionable approach to privacy. (KDE)

Don't be an incomplete and lacking project borne of frustration with other ones. (Xfce)

Re:Revert back to what worked

[couchslug]

"Don't screw up the perfectly fine UI because you have nothing else to do. (GNOME 3)"

Al UI should constantly change because change is progress.

That's why the letters of the alphabet are revised every few years.

Re:Revert back to what worked

[fnj]

Unlike, ahem, certain other DEs, each version of Xfce is markedly improved. The current version is already markedly better than even the one in Xubuntu 12.04. I can hardly wait to play with Xubuntu 12.10 and Xfce on Fedora 18.

Re:Revert back to what worked

[fnj]

If all you want is long term support, you could do what I did: adopt one of the free clones of RHEL6. RHEL6 has Gnome2.32 and will be fully supported until 2017.

Re:Best strategy

[darkfeline]

Alternative: make a cute anime girl mascot.

Re:I think Gnome 3 is lightyears ahead of Gnome 2.

[DCFusor]

I won't even argue that. But I have around 20 desktops, and zero tablets or phones. Gnone3 utterly sucks, adds nothing new, but takes away features I liked in gnome 2.

I don't need great big things wasting pixels I paid for. I don't have the first touch screen in my home. Hard to see how I could even reach most of the usual 4 23" monitor setups if they WERE touchscreens. I don't need to explore my computer on every boot - I know what's on there because I put it there.

I create things, not consume them. Why should I have to put up with a screen manage for consume-only types that really does not fit my needs and which wastes my time by removing the few features I actually do use all the time. I don't give a shit about someone saying G2 looks antiquated, because I almost never even see anything of it - I use the pixels I paid for for my apps - many of which I wrote, not to just screw around in the opsys, but you know, actually USE the damn computer to do something useful.

"Find new niches"

[mattsday]

Why does GNOME have to find new niches? It's the de-facto desktop installation for an awful lot of distributions and has been the primary choice for an awful lot of people for the past 10+ years.

It seems to me that they already had a huge user base and many more coming on-board through the likes of Fedora, Ubuntu and Linux Mint. They had a good thing going with a consistent toolkit (GTK+2), LGPL and some really nice software. From my humble perspective, this is a great starting point.

Instead they released GNOME 3. I have no idea who it's for? I remember GNOME 1.x and the thousands of configuration options - it was definitely overkill for a standard desktop environment. I think GNOME 3 is bad for exactly the opposite reasons - completely no customisation. I have no idea why they can't get this right and understand their target audience.

Fortunately, there are solid alternatives. However, I find it a great shame that GNOME seems to be determined to lose its userbase to meet some CS/HCI textbook ideal.

Re:"Find new niches"

[Dracos]

I have no idea why they can't [...] understand their target audience.

Because starting with Gnome3, they decided their target audience is tablet/touchscreen users. There has not been, nor is there ever likely to be, hardware installed with Linux+Gnome3 out of the box. They decided to cater to an audience that does not exist.

Gnome3, Unity, and the UI-formerly-known-as-Metro all suck donkey balls, assuming you don't believe the few users who have completely adapted their usage patterns and workflows, after much effort, for minimal gains. Any perceived simplicity is actually just more complexity hidden beneath the surface.

And this is all beside the fact that touch UIs are innately less capable than the traditional keyboard+pointer paradigm.

Re:"Find new niches"

[pscottdv]

I keep reading this, that Gnome 3 is for tablet computers. Where does this come from? I'll tell you where it doesn't come from: people using it on tablet computers! I tried to use it on a tablet computer. It does not work. If you ever used it on a tablet computer you would discover in the first two minute, as I have, that Gnome 3 IS UNUSEABLE ON A TABLET COMPUTER!

Gnome 2? Works fine. KDE? No problem, LXDE? Works great. Gnome 3? YOU HIT THE WALL IN TWO MINUTES! TWO MINUTES!

I actually like Gnome 3. I want to use it. I use it on my desktop and my laptop. But the Gnome developers won't fix bugs even when they are complete show-stoppers. Hey Gnome team! How about making a password dialog box that, I don't know, maybe actually allows a guy to bring up an onboard keyboard instead of taking over the desktop?

Even if they fix Gnome the Linux Desktop is a fail

[Anonymous Coward]

From wikimedia stats we see that Linux users on the desktop aren’t growing. Only Android on tablets and smartphones is doing good. Linux is stagnating at 2%the only change is about users that switch to anoter distro.
Is it important that Linux isn’t growing on the desktop?
I think it is and we can’t just say: “oh I’m fine with my OS. Who cares about the rest of the world?”. The reason is that while on the servers you can choose to use whatever software you want. For example you want to use mysql, apache, python etcfor your website? It’s fine! Do you want to deliver videos in ogg/theora format? Yes you can. Who can stop you? That is because on the server you’re the king and the users must take what you give. It’s one of the reasons why Linux had not problem to grow in popularity on the server side.
But on the desktop you (as user) don’t decide everything, because in many cases you’re just a passive actor. The Linux market share is only 2%? Well the consequences are that Adobe stops delivering the Flash Player (while before was delivering a flash player that was crap). Netflix doesn’t ship his client for Linux. Games are not made for Linux (yes I heard about Steam but we’ll see how it goes). Maybe the Olympics in your nation will be streamed using a DRM that is not available for Linux . And most important: many professional programs will never land on Linux. So not only Linux won’t attract any new users, but also this will have the consequence to cut you out from many different things that will make Linux an inferior OS choice for the Desktop.
Then some Stallman’s fan could jump out and say: but I don’t want those things! I want to stay pure and do what Stallman says: use only software that respects my freedom. Yes suretoo bad that I don’t see a lot of the Linux people using gNewSense, having no proprietary drivers installed, no proprietary codecs and watching youtube videos without using the Adobe’s flash player (probably there are better examples) . I believe that most of the Linux users are not so strict to desire a 100% open source software on their machines. They love open source, but they also don’t want to be marginalized and they care about being able to use their computer to satisfy their needs
So I said all this to explain that:
a) The small market share has side effects on users on the Desktop and so is very bad that doesn’t increase
b) Most of the people want to use Linux not because they’re crazy about Free Software, but because they want an alternative between Microsoft and Apple
c) You can’t increase the market share if you have less to offer in respect of the other operating systems

So how do you increase the market share? In my opinion: You need to make great software that is not available for Windows and OSX.
Is it possible to do that with open source software? I’ve no idea. Probably not. Also I’m sure many open source developers don’t even like it.
I think most of the Gnome developers just don’t care if Linux is at 2% of if there are some annoyances, especially because I believe most of them don’t use Linux as their primary OS. They just love working together on Gnome, but they don’t have the pressure to reach real pragmatic goals. Because that would require some compromises.
So the only way to create an alternative to Microsoft and Apple (that is what I care most) will be to hope that one day some big company creates a new brand and ships computers with Linux and at the same time makes available some of the coolest proprietary programs you’ve ever seen. That someone could only be Google. Not like Dell and HP that keeps selling hardware with Linux as a third class choice, with no marketing and no ideas behind.

Bingo

[just_a_monkey]

> find new niches. ... tablet,
> ... cloud-based services.

If only someone had said "social media" also, we'd have had the whole set.

Oh really?

[Osgeld]

You mean by fixing the standard issue list of complaints and noticing that linux nerds are NOT using their computers like large cellphones, would reduce almost 90% of complaints?

What took you so fucking long Sherlock?

Will I return to gnome even if they do what they say? I dunno ... On one hand I do like spiffy new UI's, on the other hand I dont like wasting CPU and GPU power on dumb shit like windows and special effects I never pay attention to.

Its simple

[MichaelSmith]

Just stop telling people how to work and think.

try downloading some extensions...

[sqldr]

A moveable panel

There's an extension for that..

panel applets

Many extensions do that.. it goes against what gnome say, but they work. I've got my unread mail count in my panel..

desktop launchers

Urgh.. I'm sure someone could write one. I always turn off "file manager on desktop" because having to move a window out of the way to start something is a waste of time. I normally use my desktop space with, er, windows... you can already put files on the desktop. You can turn it on with the tweak tool. KDE got it right by adding a desktop widget, so it didn't take over the entire desktop. If I want to start an app, I go "t..e..r.." ooh, a terminal in 5 key presses!

user control of virtual desktops

There's an extension for that, although once you get used to it, the "new desktop every time you use the last" option is something I really don't want to go back from. It's really efficient once you've mapped better keys to desktop switching. Especially once you have 2 monitors and you CAN'T switch desktops on the other one. It acts like a sort of main work screen while all the web/email crap is the stuff you switch. Of course, there's an app to enable switching on the other screen.

menu alternatives that would remove the need for the overview

there's an extension for that. Although i'm not sure of the "remove the need". I prefer the overview - you don't have to use the mouse in it.

all of these could be added easily as options.

They ARE options. Try http://extensions.gnome.org./ There's even a single click on/off button for each extension to turn them on and off.

Honestly, people use it for 5 minutes and suddenly think they're an expert on desktop design by saying "lets make it like gnome 2!"

Re:I think Gnome 3 is lightyears ahead of Gnome 2.

[sqldr]

I don't need great big things wasting pixels I paid for

The top bar is about 16 pixels high, and with the overview replacing the windows-95-esque taskbar, I've gained about 48 rows from the bottom of the screen. I've got more screen space than ever. I run it on 2 1080p monitors and I'm not aware of anything using the space. Besides, it would need serious work to be a tablet interface. You can move windows round and resize them, while there's no clicky thing to switch desktops. that's either keyboard shortcuts or the top corner. If you want tablet, try metro.

Car analogy

[MichaelSmith]

So out of the box every control is a switch under the instrument panel but you can install your own extensions with steering wheels, pedals, etc if you want.

Re:Car analogy

[MichaelSmith]

icons all over the desktop (which you can't see because there's windows in the way)

Why the hate? Why this insistence on telling people how to work?

Re:Car analogy

[supersloshy]

That's a bad analogy because it implies that these extensions are required to use GNOME 3. They are not. GNOME 3 is functional without any of these extensions. In fact, I only use one. Your analogy implies that, to do anything with the OS, you have to install an extension. This is completely false. It might make it easier or more comfortable for some people, yes, but it is absolutely not required.

Any recovery strategy starts with "We're sorry."

[erroneus]

The first thing that would get everyone's attention is an apology and/or acknowledgement that they did it wrong.

There was nothing wrong with wanting to create a tablet friendly UI... nothing at all. What was wrong was trying to foist it onto desktop users. Wanna make a tablet UI? Great! Do that in ADDITION to what you already had *AND* make them compatible with each other so that a user or a program can work easily in either.

The desktop isn't going away any time soon. The very notion that people are ready to move on into the tablet hype world is ridiculous.

It's understandable that no one would want to be left behind or to have a fear that you might be considered late to the party or irrelevant if you don't have one ready when the market wants it, but to push it onto the market before it wants it? What were they thinking?

And I'm sorry developers might have low morale, but that bad smell they've been wondering about isn't coming from the breath of the users complaining, it's because they had their heads up their asses... which might explain why they couldn't hear the users...

"Developer morale" should be fine since they...

[couchslug]

...are doing what they choose.

Developers don't need users so they don't need to give a fuck about what users want.

Stop hiding stuff

[SplashMyBandit]

I personally left Ubuntu for the Mac with the whole Unity debacle. For me Unity was a step backward for several reasons:

I hated having to search for applications before being able to use them. Being able to search is fine, but I found the menu structure (eg. administrator tools vs applications) in earlier Gnome was actually better at helping me find what I want than an ab initio search (which assumes I know and remember all the often-bizarre names of all the programs now on my system).

I also hated how the control panel was dumbed down to the point of being unusable. A lot of configurability that was present in Gnome 2 was removed. So when I went to change a setting I couldn't. Einstein said "As simple as possible, but no simpler". Notice how there are two parts to that sentence. The Gnome 3 crew designed by the first part of it only.

I'm a Mac user these days and I *loath* the single menu. Gnome 3 is cursed with this also. One of the things I missed when going from Gnome to Mac is that each application window could have its own menu. When you are doing stuff on two or more screens then moving back to the main screen to access the menu is a PITA. And no, I use so many programs for different purposes it is impossible to memorize all the menu commands for each application - so menu use is essential.

Reliability matters more than anything just about else. Unfortunately with Gnome 3 being new it hadn't got to a mature point where stuff works flawlessly and reliably. It's nice if the backend is "teh new shiney" and will support stuff in the future, but if you are continually reinventing the core all the time then the system never gets to be stable (plus, it takes time for applications to be built on new core tech, so every time you change the core you lose applications - and it is the applications that end users actually care about).

Just because you want to work on tablets don't forget your existing userbase. Making a better tablet workflow at the expense of smoothly working (fewer clicks) with mouse, keyboard and multi-screen is of no use to me. Hence, bye bye Gnome ol' pal.

The war with Canonical

[jbolden]

I think one of the things that often gets forgotten was that Gnome 3 ended up in a war with Canonical in March 2011. Canonical represented somewhere between 50-80% of the user base. Once Canonical came to believe that the Gnome foundation simply would not listen to their point of view and their only alternative was to fork things went downhill badly. I think its time for Gnome to admit they lost this war.

Canonical instead of pushing the advantages of Gnome 3 focused heavily on the minus. Instead of easing their customer base into Gnome 3 they moved them away from it towards their Unity / Wayland vision. Canonical could have helped to soften some of the rough edges and at the same time Gnome thought deeply about consistency and functionality issues which have haunted Canonical.

The most popular Gnome desktop is now Cinnamon which is a fork. The second most popular is Mate which is a rejection of Gnome 3 entirely. KDE developers consider Gnome to have bullied and lied to them about cooperation so Gnome is likely to see less cooperation.

There are some brilliant aspects of Gnome 3. And I could see it evolving into truly the best desktop OS around. But it won't have the time or support to do that, in the current state of alienation. They have minor technical problems but large political problems. It is time to address the politics and compromise a bit to get back to a situation where they aren't decaying rapidly.

Re:Another "GNOME 3 is bad" article, ohhh boy

[Sussurros]

Actually Gnome 3 isn't bad, Gnome 3 is a pig and now they're talking about putting a bit of lipstick on it. Gnome 2 / Mate is still around so let the Gnome devlopers kiss the Gnome 3 pig that they clearly want to so much, it doesn't mean anyone else has to.

Re:Hubris

[jbolden]

Metro makes sense financially. If it works it buys Microsoft a generation of desktop domination. If it fails, then most likely Microsoft couldn't have done anything to save consumer, falls back and spends the 2020s defending enterprise.

Gnome is in a different position.

just go back to a Gnome 2 like shell

[kenorland]

I'm sure lots of good things have happened under the covers in Gnome and the libraries etc are likely fine. The objections in Gnome 3 are mainly to the rather radical and unnecessary changes to the UI. But reverting that to something that resembles a UI people are used to shouldn't be so hard: just change the top-level graphical shell to use panels, menus, and window management in the traditional way.

for languages require IME

[causeless]

in Japan, to launch Gnome shell,

1. click "Dash" or hit Windows-key.
2. check IME is disabled.
3. Alt+Space to disable IME.
4. wait a moment.
5. double-check IME is disabled now.
6. type "Tanmatu" and hit Space.
7. check IME suggests "" ("terminal", in Japanese) properly.
8. hit Enter twice.
9. Alt-Space to disable IME.

What's a great userbility!!

There are no shortcut like Windows, type "term", Enter.
and additionaly, Japanese users must guess which translated words associated to what one want to get.
Terminal, shell, command-prompt and many other words may be translated to "". Accept both English and Japanese in launcher does not help us.

Gnome 3 is not the only thing that is sux

[Taco Cowboy]

It's the attitude of the Gnome developers

They are too arrogant

As TFA also has pointed out - they _never_ even bother to listen to the users - as if they (the developers) are "higher grade human beings" while we users are made of "lower grade materials"

That's what really sux

The "sux-ness" of Gnome 3 is but a by-product of the arrogance of the Gnome developers
 

Re:Gnome 3 is not the only thing that is sux

[bunbuntheminilop]

This isn't true. The users want a consistant HIG and design philosophy across all applications, and a consistant reliable release schedule, with an open plan. This is what they did, and it worked, for 8 years or so. Not perfectly, but it worked.

The problem is that we're terrible with the big projects. Gnome3 is just a by-product of this. GStreamer on gnome2 wasn't stable for 4 years after gnome2 introduced it in 2.02! 4 years! The problem is releasing stuff broken, and hoping the community will fix it, instead of releasing stuff working, and encouraging developers to add capability.

Re:Gnome 3 is not the only thing that is sux

[gl4ss]

I was under the impression that Gnome3 worked exactly as designed. it's just that their guidelines were either SHIT or interpreted wrongly, which is the same as being shit really.

if they want to work on movie interfaces(tm) then they can, just shouldn't expect people to use it and to donate time/money for it.. they had experts who came up with shit, that's pretty much the whole story.

it seems they've eating up the no distractions mantra to the level that it's a distraction("users are idiots"), like a bad boss.

Re:ego boost and resume padding

[Osgeld]

"Almost all serious linuxers I know have switched to lxde or xfce in desperation at the bloat and bugs"

Hell I am not even all that serious and I switched to XFCE just cause I got tired of being told that I was doing it wrong, and I was spending more time fighting the UI than doing my work.

all wrong, extras should be apps, not gnome

[cheekyboy]

All this dev is on the wrong path.

1. Window managers should be kept simple, but highly flexible, but should not contain applets/menus etc..

2. All control panel stuff, should be really part of the OS and be not tied to any window manager, but run in all of them. Just like windows, can code for win32/.net/wpf/metro, just like the main linux UI api, aka gtk or qt. A WM should not be tied to those two. But perhaps have a higher level abstracted api that can use either. Apps/Applets can communicate to other apps or the WM via the DBUS, or via a core common api that is not tired to a WM.

2. It sounds like both Gnome&KDE need to work together to create a new layer thats common to all linux's, perhaps like a linux core desktop layer spec. XCORE perhaps. And their cute custom Wmanagers can sit on top, where a commonly written control panel system (part of XCORE) can run on both WMs.

3. Linux needs to redesign how X + WM + GNU work together. X11 + XCORE(qt+gtk+scriptbased api) + WM on top.

This way, the WMs can be more like 'theme styles' with applets.

Any way , too late, nothing will happen, and googles Chrome Desktop OS in JScript/Dart might take over, or some sort of hybrid Android 5 GUI with full desktop features might end up killing both GNOME/KDE if it + ADK can run inside any linux.

Re:What theory is it?

[afgam28]

They have a page on their wiki, which explains each UI element and why it is there.

https://live.gnome.org/GnomeShell/Design/

Unfortunately it focuses on what they put in, and doesn't really explain much about why they removed or changed a lot of things that people are used to. Here's some of the reasoning behind the more controversial things:

Why the dock/taskbar was replaced with the overview:
The overview is much better than a taskbar or dock, when you think about it in terms of Fitts's law. Targets should be big, and the Expose-style overview provides bigger targets (window thumbnails) than either a dock or a toolbar. Also the hot corner is one of the five fastest targets that a user can hit (the four corners of the screen and the pixel below the current mouse position).
One thing that I've noticed is that users who complain about task switching don't understand how to take advantage of Fitts's law. All you need to do is learn to flick your mouse quickly into the top-left corner. Don't bother aiming, the edges of the screen will guide your mouse into the corner.

Why alt-tab seems broken:
Windows uses a "window-centric" whereas Gnome Shell uses what is called an "application-centric" model.
The way Gnome Shell works is quite simple: Alt-Tab changes between applications, and Alt-` (the key above the tab) switches between windows of the same application. It's exactly how Mac OS implements it, but lots of people hate it because it breaks the habits they learned when Windows 95 came out.
If you think about it, tabbed applications (like Chrome, Firefox, and lots of Gnome apps like Terminal and Gedit) don't really need to have tabs. The tabs are just there because window-centric desktop environment doesn't provide an easy way to switch between windows of the same app. Just press Alt-` and switch between, say, different terminals, without worrying about mixing them up with web browser windows.

Why there is no shutdown menu option:
Your computer has a power button, and you can just press that. It should initiate an ACPI shutdown, and it is by far the most obvious way to turn off your computer. But people disagree because years of using Windows (or one of its clones) has conditioned them into thinking that it is totally intuitive to navigate through menus and select a shutdown option instead (if you really want to do this, you can install the "alternative status menu" extension).

Why they removed the maximize/minimize buttons:
https://mail.gnome.org/archives/gnome-shell-list/2011-February/msg00192.html
What's important here is that they didn't remove the ability to minimize or maximize, but rather just the window manager buttons. You can still do it using other methods. In particular, most people maximize by double-clicking the title-bar anyway.

You're right that the UI would make sense to people if it were explained better. I really like Gnome Shell but the biggest failing is that they didn't communicate their vision clearly, and didn't provide a smooth transition from Gnome 2 to Gnome 3. They just said "here it is, and here's a crippled fallback if you really disagree with what we did".

I hope MATE with GTK3 ends up being the transitional environment that the Gnome people should have provided. And I also hope that more users give Gnome 3 a try with an open mind, because it's a really great desktop once you get used to it!

Thoughts from a core GNOME 1.x and 2.x developer

[Anonymous Coward]

I'm an old core GNOME developer, around for the 1.4 - 2.x days. I haven't been involved in GNOME 3, but I think they're on to some really cool things, even if there are serious problems now. These flamewars make me sad.

Many (most) of these comments remind me of the same slashdot.org discussions between GNOME 1.x and GNOME 2... I should remember; I was one of the core GNOME 2 devs who was flamed to hell.

Now people are talking like GNOME 2 was some sort of epitome of Linux desktops, and couldn't-we-just-stick-to-that-pretty-please. It also reminded me of the flack that KDE 3 developers took. Talk about whiplash. I don't think many people comparing GNOME 1.4 to GNOME 2.32 would prefer the former, and yet, to hear the cries on slahdot at the time, GNOME 2.x was doomed and nobody used it, and nobody would ever use it. Dooooooooomed. Doooooooooomed I say.... because we were all such complete idiots that we couldn't tie our shoelaces without shitting our pants. ;-)

I notice two things:
1) Free software desktops are often a little half-baked between major UI revisions. This does suck, but I think its a outcome of volunteer hackers... sometimes its hard to wait long enough to add all the features people like and miss before doing a major rev. Frankly, an effect you often see is a decrease in hacking if a project goes too long without a release (makes sense psychologically, right? sort of related to delayed gratification....).

For example: GNOME 2.0 was stinky. People flamed the hell out of us (in many ways, rightfully, it was half-baked), and not JUST about our current state, but speculatively that this represented some insane mis-step for the project. Instead of imagining what the negative-changes could allow in the future, they pretended like we were retarded, and driving the ship as fast as possible straight to hell. No benefit of the doubt. Now I don't want to apologize for this, I think free software should be held to the high quality standards of commercial software, but I mention this because its important context to the sort of panic-reaction people are displaying, assuming GNOME 3.0 betrays some fundamentally flawed direction rather than viewing it as "released too early, too half-baked, before certain necessary things happened".

By GNOME 2.6 it was pretty awesome. By GNOME 2.12 pretty much everyone just shut the fuck up. A number of users found GNOME 1.x more to their liking and moved on to other desktops, but we picked up Waaaaaaaaaaaaay more users than we lost. Today, I think most people would cringe if they had to use GNOME 1.4 instead of GNOME 2.12 (or whatever).

So: GIVE GNOME3 SOME TIME, and view GNOME releases with a fresh eye. GNOME 3.8 might rock your world, and the 6-mo release cycle means changes happen faster.

2) I think if you asked the average slashdot reader, they would like to think they are more "open to change" than the average citizen. In fact, I find the entire *nix culture extremely resistant to change, automatically viewing change they don't understand as "change for change's sake". In a way, its sort of unique and cool.... most of the western world is swept up in a progressivist notion of time, always viewing the future as "better" than the best. In contrast, *nix culture often has a distinct note of Indian-style views of time: the gods used to walk the earth, and since then, its mostly been decay. The downside is that its not a very fun community to develop UIs for: instead of focusing on "what's gained", people pull out flamethrowers immediately at the slightest hint of something being lost. CHANGE USUALLY REQUIRES LOSS because DESIGN IS BALANCE. Sometimes the balance is wrong, and sometimes tradeoffs are made when they needn't have been. I think just like GNOME 1.x to GNOME 2, sometimes the first-couple-passes you lose more than you needed to, and this gets balanced out over time.

As a bystander to GNOME 3, I see many ways they could achieve their goals while minimizing the (very real) losses hackers are experiencing whe

We have

[Rix]

Which is why we've all moved on to either other DMs, or forks like MATE or Cinnamon.

Re:I think Gnome 3 is lightyears ahead of Gnome 2.

[afgam28]

I remember a few years ago when my dad started using Ubuntu. He'd previously used Windows all his life but was sick of all the spyware on his computer.

At one point he called me and said "all my windows have disappeared!" Once I saw what he'd done, it was obvious - he'd changed workspaces and his all windows were on the previous workspace. But he had no mental model of how workspaces worked, and he wasn't even sure if his documents were still open. When I fixed it for him, he remarked something about Linux being really complicated.

When I installed Compiz and enabled the Desktop Cube animation, he mentioned that workspaces suddenly made sense. If he accidently switched to the workspace on the right, it was obvious how to "fix" it - you just need to rotate the cube back in the reverse direction.

Sure, it's eye candy to us, but animations can be used to help users understand what is going on in a desktop. Most Slashdotters are probably familiar enough with workspaces that they don't need to think about them, but keep in mind that it is a completely abstract concept. Animations can help communicate to new users how UI elements have been, and can be, manipulated.

Re-architecting the layers....

[unixisc]

2. It sounds like both Gnome&KDE need to work together to create a new layer thats common to all linux's, perhaps like a linux core desktop layer spec. XCORE perhaps. And their cute custom Wmanagers can sit on top, where a commonly written control panel system (part of XCORE) can run on both WMs.

In that case, it could become a part of Wayland, so that all DEs can benefit from any Control Panel/Configuration settings. Or if it is a part of the underlying OS, it can be something in Wayland that enables the WM to make that tool available to users. That way, it will work the same no matter what the DE.

3. Linux needs to redesign how X + WM + GNU work together. X11 + XCORE(qt+gtk+scriptbased api) + WM on top.

This way, the WMs can be more like 'theme styles' with applets.

Any way , too late, nothing will happen, and googles Chrome Desktop OS in JScript/Dart might take over, or some sort of hybrid Android 5 GUI with full desktop features might end up killing both GNOME/KDE if it + ADK can run inside any linux.

You seem to be suggesting for Qt/GTK/... to be a part of Wayland, which currently is just OpenGL plus some basic compositing functionality. I'd think that adding another layer on top of X would be just increasing the bloat, which is another complaint that people have about Gnome3, Unity and so on. I agree that the roles of who does what need to be better demarcated, but beyond that, adding more layers or standards is just going to confound the issue.

As an example, if the OS already offers a way to configure a network, why should KDE or GNOME come out w/ something new, like kwlan? These sort of functions, which are handled by the underlying OS, should be there in a graphical interface, but in one common to the DEs, which is why I was thinking that it would be a good thing to have in X or Wayland (preferrably the latter, since who knows what changes will have to happen to X) But for the individual DE functions, I don't think that Qt or GTK libraries need to be a part of Wayland, particularly since you now have different DEs based on different versions of the same libraries i.e. Mate based on GTK2 while Cinnamon on GTK3, KDE on Qt4 while Trinity on Qt3, and so on. If they are to support all versions, that would create needless bloat. So just have the DE bring w/ it whatever library it uses, and don't include those libraries w/ the base userland install.

Re:All those DEs were "borne out of frustration"

[unixisc]

GNOME was born not so much out of frustration w/ KDE, but rather, a part of those license jihads that RMS likes to wage. GNOME was born b'cos Qt at the time was licensed under the QPL. Once it became dual licensed, the reason to have a GNOME was no longer there. If they really wanted to do a GNU Networked Object Model Environment, they should have taken GNUSTEP, studied it and modified it to suit whatever their idea of a networked OO interface was.

Given that currently, none of the DEs are GPL3, the GNOME team could just make it that, so that the 'Libre-Linux' crowd can add one more GPL3 piece to the puzzle. KDE will remain LGPL, Etoille will remain BSD and others will go w/ whatever licenses they are comfortable w/.

What do you mean assumtions?

[dutchwhizzman]

First of all, skimming forums for feedback about the changes in Gnome3 gives you zero people that appear coherent throughout their posts that actually like the changes, apart from some Gnome3 developers. Go figure. The amount of people bitching about not being able to do things window managers have given people since TWM and CDE were the latest thing is simply overwhelming.

Second of all, tablets may be getting more popular, but you're replacing desktop user interfaces so at the very least, retain the features, possibly configurable, that make up a decent desktop window manager. For instance, no screen saver configuration or selection? What?? No hot corner selection? You need third party plugins to get you an icon you can click once to open applications?

You may be right about making assumptions, but it's not this guys task to do research in to what users want and how they like the changes. That task is for the gnome development team and they haven't done that ever. Not before, not during and not after the release of Gnome3.

Now what case can be made for gnome3 changes? I haven't seen one tablet manufacturer that adapted Gnome3 as their UI, I've seen literally hundreds of users complain, I haven't seen more than a handful people that like the changes, most of them being Gnome3 developers and thus biased. If you want a case to be made for the Gnome3 changes, why don't you do so yourself instead of blaming other people they're not doing it for you? What are those merits you are talking about? How much users has "gnome" gained since the introduction of Gnome3? I'm willing to bet the absolute number user base has dropped, while both Win7 and OSX have grown, so comparing Gnome3 to those makes Gnome3 look bad.

Re:Thoughts from a core GNOME 1.x and 2.x develope

[tajribah]

The core of the problem is that GNOME developers have the habit of releasing as 2.0 or 3.0 something, which is of beta quality at best. It's quite possible that GNOME 3 contains some great ideas, but trying to attract users to software, which will need a year or two more to reach usability of the previous version, is not going to win anybody's sympathies. Exactly this has already happened with the release of GNOME 2.0: its usability was nowhere near that of GNOME 1.x, but still, it was presented as a replacement of 1.x. The users were rightfully complaining. One would have hoped that GNOME developers have learned something from that fiasco...

As of culture resistant to changes: For most people, the computer is a tool. And as with many complex tools, it takes time (sometimes years) to learn how to use them in the most efficient way. The learned experience is very valuable, but a part of it is necessarily lost when the tool suddenly starts behaving differently (people are not used to their screwdrivers changing shape overnight). Sure, changes are necessary for progress, but you should not ignore that changes come with a high cost to the users and radical changes of basic concepts even more so. Changing details is usually fine, removing functionality is worse, and radical changes of established products should be done only in cases, where the benefit is an order of magnitude larger than the loss. GNOME developers seem to ignore this fact of life for years.

Article fail

[allo]

> A moveable panel, panel applets, desktop launchers, user control of virtual desktops, menu alternatives that would remove the need for the overview -- all of these could be added easily as options.
most of the listed points would be more or less major changes, nothing easily added.

The best and easiest thing would be to start again working on gnome2, releasing a gnome4 which is based on 2.

Re:What theory is it?

[epyT-R]

1. the thing with fitt's law is that its relevance drops off over time, leveling off once the user gets more and more familiar with the environment and his mouse settings. At this point, huge targets and fullscreen menus hinder usability and workflow, and the taskbar concept becomes more efficient at starting and managing application focus, takes less space, and does not steal visual focus from what's up on the screen If you have multiple monitors, flicking mice into corners isn't a given either. This is true of windows' start menu as well (which by the way lets you start applications with no fuss, without typing, is customizable, and again does not steal visual focus from what is already running on your screens).

2. There's something to be said for the ability to sort through all open applications, then by their windows, but the key lies in how they're sorted. However, browsers have tabs for a reason.. I like the fact I can minimize the browser and all my tabs stay with it. If I need a separate window I can detach or open a new one... I remember the pretab days when I had 4000 browser windows open.. that sucked.

3. sigh.. if having a suspend menu option is acceptable, then so is having a shutdown/restart and poweroff. In different situations I want to do differnt things. The menu should allow this, with no fuss. The power button on the computer is only useful for triggering a default option (usually suspend by default, thus duplicating a feature when your whole argument (or theirs) is one of deduplication). if you are lucky, you can set that in the cmos. On most OEM equipment, you can't. Not including this is fucking stupid. Period.

4. You're right that most double click to maximize.. However, then why include that useless drag to top to maximize feature that just gets in the way when the user just wants a window at the top? The drag left and right is annoying too because sometimes I want the window open but off the edge due to insufficient room. It just depends, but certainly the last thing I want the window to do is snap back to the screen edge and take up half the display! that's in fact the opposite of what I want. That shit is on windows too and it sucks. At least I can shut it off on that os. How about a snap to edge/window feature..with a toggle to turn it off if it gives problems? Most window managers do this already.

They could try reverting to Unix's origins

[ras]

When I started with Unix too long ago, the philosophy was to develop a tool box that let people customise the box to their liking. FVWM was like that. It was flexible enough to turn something that looked like Windows 95 and OSX, but configuration was not for the feint of heart.

What I expected to follow FVWM was something even more flexible, that solved the configuration file from hell problem. Something so flexible that you could have emulated the current Android, Windows 8, and OSX interfaces. So, for example, you could decide whether you wanted the applications menu bar to be per part of each window like Win 95, or have a single global one like OSX, or not display it until a button press like Android. Where you could decide whether you wanted to run every application full screen, or in its own window, or something in between split over multiple virtual desktops. FVWM already came with a collection of menu, dialogue, and panel and task manager widgets - I expected this to be expanded in the usual open source way so there were 1000's of them, most of them useless, but spurred on by the toolbox mentality that made experimentation with news ideas was easy. I expected to a fight between API's that allowed you to play with touch and multi-touch, so that someone could in principle make an upward five fingered swipe with pike launch vim, or a three fingered back swipe would invoke the browsers back function, or a two fingered circle would do an Alt-Tab, the direction depending on the direction of the circle.

But that is not what we got. In fact, the reverse happened. As others have pointed out, instead of making Gnome 1.0 more flexible, the Gnome developers decided to solve Gnome's configuration problems by removing most of the configuration. In Gnome 3 it got to the point that when I decided the fonts Gnome were using were too big even the ability to edit that was taken away. (You have to install some tweak program.) Worse than that, not only can you not configure the layout and behaviour of Gnome 3 for the platform you are working on, it seems to be optimised for a platform no one uses it on - a small screen with a keyboard ?!?!

Look boys, I'll spell it out for you - the world is NOT converging to one platform everyone uses. The reverse is happening - it is splintering. Whereas a few years ago you could safely assume all your users has a large screen, mouse and keyboard, that assumption makes no sense today. At one end people have 3 x 27" 2880x1440 screen hooked to a single computer with a mouse, keyboard and stylus input device. And then we smoothly move through a number of form factors end up at some 3" screen has a touch screen without a keyboard. In this world you can not dream up the one true interface and expect everybody to be happy with it. The very idea is insane.

In the world we have today, the Unix way of providing toolbox people can use to customise to their environment should be having it its time in the sun. Sure it's more complex than iOS, but unlike iOS it could be made to work on everything, and unlike iOS we don't have to cater to unsophisticated users. Instead we it seems we have lost the original Unix philosophy we started with, and the result is rejection by the only people who used Linux on the desktop - the people who were attracted by that philosophy.

I didn't consider myself a normal user—

[aussersterne]

I submitted detailed bug reports and did a lot of repository code testing for KDE, and submitted code myself during the 1.x and 2.x series to a few parts of KDE. No, I wasn't a major contributor or developer—I had/have a real job and it wasn't KDE—but I considered myself just another tech-literate community member that could help out with a snippet of code here or a bug fix there.

I anticipated doing the same with 4.0.

But there's a level of "not ready yet" at which you can't even operate at the level of "if I see a bug I'll try to fix it and send it along" because you can't even stay logged in long enough to encounter a single thing that could be isolated as a "bug." KDE 4.0 was released at the "general framework is still under heavy development and really doesn't even work level," the level at which the things that go wrong aren't about bug reports or about bug fixing but are at the level of deep familiarity with the evolving codebase in large chunks.

They shouldn't have said that KDE 4.0 was a test release not ready for prime time. They should have said it was alpha-level code that was more or less guaranteed to break early and often, and should be installed only by those willing to help bootstrap a codebase that was just to the point equivalent to "kernel now boots on the device but no userland yet and/or highly unpredictable/unstable userland guaranteed to come crashing down within 10 minutes of boot." Because that's basically where KDE 4.0 was as a graphical environment. It wasn't ready for beta testing or development of apps. It was only ready for people that wanted to seriously bang on core KDE.

That would have accurately described what they released.