Linux vs. Windows: Choice vs. Usability

ThaReetLad writes "In this article at DevX, Executive Editor A. Russell Jones makes the case for a standardised GUI for Linux. He argues that the promotion of choice of GUI as a positive feature of using Linux is detrimental to its chances of attacking Microsoft's home user monopoly. From the article: '...the open source community must recognize that its primary goals: freedom of choice, freedom of source code, and freedom to alter applications, are not the goals of the average user.' In particular he argues that the choice of desktop between KDE, Gnome, IceWM etc, is not one that a former windows user, even a fairly technically competent one, is going to able to make an informed choice on, and that they should not be forced to make that choice in order to get good use out of any applications they might want to use."

Comments

Good idea

[mao che minh]

The advancement of KDE and Gnome can occur beneficially without the standardization of either. Despite this, it would be much to the benefit of all Linux companies if they all worked together on one standard desktop, instead of leaving it up to the community. The Open Source community will continue to make new GUIs and make the existing ones better, but Linux companies should be interested in making the best possible operating system for both the server and the work station (what sells).

The server side of things is coming along nicely. The work station side is severely behind the competition, and the reason is directly linked to the failure of all parties to strategically target the GUI togther, instead of independantly using different GUIs that are never really that much better than another on any given Sunday.

If I had to choose, I would vote for KDE.

Re:Good idea

[DrXym]

It would be beneficial to both KDE & GNOME if they got together and sorted out a single standard things that both have in common.

• A theme engine for rendering widgets, scrollbars etc. XP & OS X have this and it's
• Icons for toolbars
• Metrics
• Usability guidelines
• Menus and icon properties

At the end of the day, most people don't give a toss what the name of the thing running their desktop is, or why KDE is better than GNOME or vice versa. They just want a consistent desktop and a consistent set of apps running on top of it.

Of the distributions so far, Red Hat has clearly gotten the message. The RH9 desktop with bluecurve theme throughout is a wonderfully put together desktop. It's only when you contrast it with the slapdash Mandrake desktop for example that you appreciate the difference that consistency makes.

Re:Good idea

[bigjocker]

This WorkStation thing is a big problem that can (and currently is) hold back more adoption of Linux in the desktop.

We have reached a state where almost every WorkStation job can be performed efficiently on Linux, with a great selection of apps and tools that compare and in a lot of cases surpases their Win/Mac counterparts, but (unfortunately) the users have been taught to rely a lot more in the OS/Desktop than it should be, instead on the applications that implement the functionality.

If you ask Joe User "what is windows?" he will start talking about the task bar, the Start Menu and a lot of images that user has fixed in his mind. If you try to push linux on them, you must have a familiar look that they can be used to, even when they sit in a different computer. I'm not talking about us geeks but the everyday users that ultimately stack up to give Windows the 95% (or so) in the Desktop market.

I have said it before and will continue to say it, Linux has been ready for a long time for the Desktop, the applications are ready and even a lot of users are ready to use it, but the main goal of Linux is also holding it back: choice.

It's a complicated issue, the freedom of choice is what got us to use Linux, and I love to use gFTP and Evolution on KDE, but that freedom of choice is also scaring potential users away.

My vote goes to KDE too ;)

Re:Good idea

[anthonyrcalgary]

I would vote for KDE too.

I think the article is stupid, but Apple has won a LOT of support through standardization. Not only of what they look like, but how they're laid out. They've got very, very specific GUI guidelines, and that's a Very Good Thing. They're purely voluntary, but they create a consistancy between software included with the OS and 3rd party software that is unmatched elsewhere.

If KDE (or whoever) were to come out with similar guidelines and most people were to follow them, Linux would benefit. Every non-OS X UNIX would benefit.

Re:Good idea

[AntiOrganic]

KDE has Human Interface Guidelines already, and they're really completely nondescript and aren't very cohesive at all. Gnome 2's are really much better and even match Apple's fairly well, in my opinion.

Re:Good idea

[Lodragandraoidh]

I Don't think its a 'dilemma' because the open source / free software developers will never standardize the GUI interfaces - precisely because people have come to like particulars of their given environment, or want the flexibility to change.

Commercial Linux packaging companies might decide to create a 'standard' GUI - but that will not slow down current and future interfaces from continuing to be developed.

Re:Good idea

[DrXym]

I have Mandrake 9.1 running on my laptop and RH9 on my desktop. Mandrake 9.1 is a dogs breakfast by comparison. I appreciate it for other reasons, but Red Hat beats it to death for general polish and then some.

The issue is that they seem content to slap together a generic KDE with a bunch of homegrown tools and think that's enough. Consequently, you have a confusing clutter of apps in the menus with names that are Kompletely Kryptic and Konfusing, a "Mandrake Control Center" vs a "Control Center" (wtf?), duplicate apps that seem to do the same thing (e.g. two floppy disk formatters, multiple RPM management apps), a help system which offers help on KDE but not Mandrake, "Drak" tools that ask you if you want to apply the changes when you haven't done anything, and numerous other silly and slapdash faults.

By comparison, RH9 is a beautiful and simple desktop. BTW I'm not comparing the underlying KDE & GNOME here, I'm comparing the amount of effort that has gone in afterwards to clean and polish things up.

Re:Good idea

[CaptnMArk]

I am the author of icewm, one of the choices :)

For one choice only, I would choose mozilla ie. the XUL toolkit.

Why? Because XUL is XML and not directly dependant on C or C++ bindings.

Because I simply don't see the problem of C vs C++ toolkit choice resolving itself any time soon. And there are other choices too, like Java Swing, and whatever mess they will do with C#/WinForms/Wine.

What we really need is a more abstracted core desktop, more document oriented (and more like www.nakedobjects.org but based on XML interfaces) and less app centric. This will allow more decoupling between GUI and actual applications.

Re:Good idea

[arivanov]

Blah, blah, blah, blah using different GUIs blah, blah, blah.

Well, a Windows user is capable of making a choice. Do not consider a windows user to be a moron by definition as he/she aint.

I have several ex-windows users around me which are by all means linux users now. When I say by all means they are using it. They are not administering it, configuring it, tweaking it or in any other way wasting their time. They actually use the machine for work.

And guess what they use - good ole Windowmaker with the standard brushsteel Debian theme (yes they have tried Gnome, Kde, XFCe, whateverE and they hated every moment of it). After all, people severely understimate the extent to which people like their machine being fast (even when it is a PIV at 2.4) with half a gig of RAM.

Re:Good idea

[NaugaHunter]

but (unfortunately) the users have been taught to rely a lot more in the OS/Desktop than it should be, instead on the applications that implement the functionality.

The human interface is far more important that the implementation behind it, and this is universal across all technology. Think of cars. If the steering wheel, radio, pedals, etc. are all in accessible places, how many people really care how they work the car? Or the phone - how many different keypads are there? Have you ever seen a light switch on a wall that went left-right and not up-down? There are a number of 'better' keyboard layouts - has anyone not included QWERTY and survived?

The human brain functions on pattern recognition. A consistant interface allows the brain to function at a 'higher' level. Let me explain - you don't think about how to walk, chew, or scratch an itch, right? You just think "I want to go there" or "Mmmm" or, usually, nothing - I sometimes don't even really I have an itch as the brain as already recognized that it should be scratched and done so before the message reached my conciousness. A consistant GUI allows a user to think "Close this window." An inconsistant GUI forces the user to think "To close this window, I need to find the X. It was on the right last time, but that's a Circle. Oh, there it is on the left. Click."

Having to think about every single action is very frustrating, whether it's a new GUI, a new video game, or even a new VCR or Microwave. And this is exactly what people have to deal with in switching to Linux - learning a new set of patterns, which requires think about them at the beginning. It's been my experience that average people barely understand the computers they currently use - they were taught 'Click here, then here, then here'. It's virtually impossible to teach them a new interface if they never knew why they were clicking in the old one.

Re:Good idea

[ichimunki]

Well, there's no reason that distribution vendors can't do as Red Hat has done and essentially do their best to merge the two major desktops in terms of look/feel and default appliations. Personally I don't give a rat's hindquarters what will get Linux a bigger "share" of the desktop. This is the flawed thinking of the open source movement, that somehow things like "market share" are measures of success.

Not true. Absolute freedom in the software is what's important. As long as we have that, any GNU/Linux vendor in the world is free to cooperate with other vendors and standardize or not standardize as they see fit. A properly functioning economy (both in terms of money and ideas) requires lots of small companies (suppliers) all competing for the dollars/yen/euros/attention of the consumer.

At this time, however, the majority of GNU/Linux users (fortunately) value freedom and choice. Any move to "dumb down" the GUI or make it more rigid without providing a big red button that says "click here if you want to take back complete control of your system" will simply alienate the current user base (that would be those of us who contribute dollars, time, or money to groups like the FSF, Debian, KDE, or even SUSE or RedHat in the form of buying the box sets).

Personally I feel like users expect too little from their systems, and too little from themselves as well. Computers are complicated machines that require a certain level of know-how. If the user wants simple, I don't see how they can go wrong with existing distributions like Red Hat.

The idea that we should "start with a standard GUI" (to quote the article) is nonsense. What we need is to demonstrate value to users, both in terms of the value of freedom (which is pooh-poohed in this article, I think) and in terms of the excess funds spent on proprietary software that users are only using minimally (you know they hit the main 10% of the features, but the other 90% they paid for go to waste). Plus, the real need here is to get preinstalled systems out there at affordable prices and offer some decent support and training options. If people can't walk into Circuit City, Fry's, or Best Buy and pick up a GNU/Linux system, forget it. Look how well Apple does against that, and they have extremely polished, monolithic systems with some of the best ad campaigns known to man on their side.

Re:Good idea

[Skeezix]

If I had to choose, I would vote for KDE.

And I would vote for GNOME. You see, that's the problem. Neither project is going to give up their work. Neither platform is going away. I know for a fact the GNOME developers are very proud of their work and are convinced that GNOME is the desktop of the future, but I'm sure the KDE developers feel the same way.

Re:Good idea

[jtev]

Damned straight, we don't need to bow to the pressure to use a sandardised anything, we need to improve our marketing and SELL CHOICE. I don't care if the unwashed masses are put off by choice, give them a good default, make choises easy to find, but also easy to avoid. I personaly love GNOME, but KDE has good points to, do I think any less of a freind of mine who likes KDE? No, but I do love GNOME. I also loved Enlightenment 0.16, and when the E folks come out with their next version I'll be rather willing to try it out. Think of it this way, in everything but computers consumers love choice they like to choose their cars, their airlines, their greasy burgers, after all shouldn't we just standardise on one fast foot resturaunt, then you could get McDonald's fries, Wendy's frosties, Burger King's hamburgers, and Sonic's limeade all at one place, but that doesn't happen, you get 4 different resturaunts that you can choose which food element is most important to you, and go to and get passable choises for the rest of the items consumers understand choice, and they aren't afraid of it. They simply need to be educated about the choices out there.

Re:Good idea

[ILEoo]

There's this thing called freedesktop that defines common standarts on Drag and Drop etc. Gnome & Kde are starting to follow it. and even xfce 4 is compatible,so it's even easier to cross use things

Re:Good idea

[BrokenHalo]

The whole point of the issue is that we do not want a "standard" rammed down our throats.

Microsoft has had years' worth of value in promoting their products as so-called "standards", and many of us have had enough.

If "someone" were to decide that the "one true" interface was going to be KDE or Gnome or whatever, many happy users of open source software would object, and rightly so.

If that means that Linux will not overtake Windows on the desktop market, then so be it. Let's face it, most Windows users will not even consider switching anyway.

The whole point of the open-source initiatives is to allow the user to make choices. Most (or at least many) users want to be told what to do, not to make choices as to how to do it.

Re:Good idea

[Phillup]

Well... in one way they have standardized.

They run on X.

So... perhaps the answer is to rework X.

I personally think that I should be able to use which ever interface I want... but, there is certainly some core functionality that could be abstracted out to a common layer, IMHO.

Perhaps this would be a good way to start? Standardize some of the functionality that is common... and perhaps provide a "Common Interface" API for programmers.

I personally think of Gnome and KDE as toolkits for application development... and not as a Desktop GUI, mainly because I can run Gnome apps on my KDE desktop... and vice versa.

Re:Good idea

[BigBlockMopar]

I Don't think its a 'dilemma' because the open source / free software developers will never standardize the GUI interfaces - precisely because people have come to like particulars of their given environment, or want the flexibility to change.

In a sense, precisely the flexibility and diversity of console applications and daemons which makes *nix ideally suited to server duty, is exactly what makes its mass-adoption on the desktop difficult.

On one hand, it's great that there might be 50 different ways to implement cut-and-paste functionality, but the end user doesn't care as long as he can cut something from one application (say, OpenOffice) and paste it into another (say, Gnumeric). On my Linux workstation, that's tough.

Never mind the lack of OLE support. Stuff that Windows users take for granted - Windows Media Player embedding in a PowerPoint presentation - has been around since Office 95, and yet I've never been able to get xine or mplayer to launch from an Impress presentation.

These are the hurdles which will need actual in-stone standards to overcome, and some of them will probably require work right down to the very architecture of the window manager in order to implement.

And we have to overcome them, in order to make Linux a viable alternative operating system for any casual user who does more than read e-mail.

I have a rant on the subject. It's a bit dated now, but trying to work as a "regular" user in Linux - without writing shell scripts or firing up gcc - I can't get done what I need to get done. We don't need more eye candy or stuff like that. We need decent apps that don't feel like the works in progress they admittedly are and don't silently close if you run out of disk space. We need volunteer UI designers to walk into offices and seniors homes to find out how we can make a more user-friendly and consistent UI without alienating the power user.

And we need to support at least every UI feature of Windows 2000 and its central apps. Let's think about it - in a field which moves as fast as IT, I'm suggesting we should use the OLE capabilities and consistency of a three-year-old product as a role model. Why? Because there are still things which Windows 95 users can do in three mouseclicks that would require firing up gcc.

Like it or not, every end user is going to compare KDE/Gnome/whatever to their experience with Windows. They don't care that the Linux kernel never crashes if the KDE crash handler is popping up every 5 minutes or if applications lack features they require.

Think of the end user with every single line of code that you write. Take a picture of a technically inept parent or aunt and stick it on the side of your monitor to remind yourself.

So what's the story here?

[Tyler Eaves]

Most of the newb friendly distros through one of KDE/Gnome in as the default choice, which works fine for someone who doesn't know any better.

Windows suffers same problem

[Doesn't_Comment_Code]

Windows has the same problem. The Win98 desktop is NOTHING like the XP desktop. Each edition they release is a little different in terms of menu placement, control panels, what's where... The only advantage is that they release one at a time, so there is only one current OS. But to go between Windows machines, you still have to adjust and know what you're doing.

Down to distribution

[cjcormack]

Surely this is the job of the distributers not the developers.

People just don't get it sometimes.

[thud2000]

How many times have Linus or others said that the goal for Linux is NOT to attack Microsoft's monopoly, but simply to provide a freely usable and stable UNIX-like operating system for anyone who wants it. These analysts can't seem to wrap their minds around the fact that "Linux" is not just another company out to rule the desktop.

Re:People just don't get it sometimes.

[v_1matst]

I can't agree with this more. Those of us who want Linux on our desktop have it on our desktop, those who don't... well... don't. I do not understand this mission to have linux as a "valid" desktop operating system. People who use it know it works (quite well in fact) and find that it suits their needs. To Joe Blow user, Windows might very well suit their needs and they find no need to go to some other system just because it isn't a microsoft product. There are people out there who have used Linux/UNIX variants and say "hey, that's great... I'll stick with windows" and I am having trouble finding anything wrong with that.

Re:People just don't get it sometimes.

[KevinIsOwn]

Exactly.

As this guy said the goals of the open source community are " freedom of choice, freedom of source code, and freedom to alter applications" and if they aren't goals of the average user that's a tough break. If they can't use one of the desktop environments like KDE or Gnome now why would they be able to use a "standardized one?"

Maybe the best solution to the whole "average user" problem is to make a dumbed down KDE/Gnome that are "easier" to use (although I don't think they are really all that hard to use as it is, and things like lindows make linux even more useable)

Re:People just don't get it sometimes.

[forsetti]

That is true -- Linus/Linux is not out to attack the MS monopoly. But RedHat, Mandrake, Suse, $FAVORITE_DISTRO are. RedHat, for example, has already recognized this issue, and started attacking it with 'BlueCurve'.

Re:People just don't get it sometimes.

[quinine]

Then the onus is on the company doing the embracing to set the standards. This is no different than in a Windows environment, where an organization must choose its applications. Though many businesses are content to standardize on Microsoft entirely, there are many that pick and choose to create a standard work environment.

There's nothing wrong with a company deciding that all Linux desktops will be Red Hat running KDE, but if I go home and install Gentoo with GNOME they can't even try to tell me I can't do that.

Consumers do not want choices...

[Anita Coney]

My father-in-law worked as a travel agent at one time. He said travel agents never give more than three choices to a client. If you gave them more, they'd have to go home and think about it.

People don't like making choices, it takes away time, energy, and they risk being wrong. That's one thing Windows (and Apple) does well, all choices are made for you.

The problem I have with the post is that it does NOT have to be a zero-sum game. If someone wants to make a distro of linux that provides limited choices, what's stopping them? Why does every distro have to be limited in choices. That mentality makes no sense.

Re:Talking head moron

[Epistax]

We have a standard desktop? Not with linux you sure as hell don't. Is KDE the standard desktop? (flamewar insues) is CDE? (flamewar insues). What are the advantages of emacs and vi? (flamewar insues).

If you can't tell from this why someone who doesn't like geeky things (aka average computer user) is put off by linux...

On a side note I think would be rather nice of distros of *nix and gui's and etc. would specify what they think they should be used for. A given windows distro explicitly states what it is for: Small business, server, home use, hand held, etc. On linux? You've get twenty distros all trying to do everything. Give these people something to grasp!

A thought

[rknop]

Whether or not I agree with the conclusions, for the time being let's accept them for the sake of argument.

Suppose that the current goals of the open source community (freedom, choice, etc.) are inconsistent with GNU/Linux taking over the desktop.

Do we then really want to take over the desktop?

If we have to become like Microsoft to defeat Microsoft, then what's the point? *If* we were just another proprietary software company, then, yeah, sure, that's the right thing to do. Since, after all, the ultimate goal of any company is just profit. The open source community is very different. The community isn't going to get rich and retire. They're mostly in it because they like the software and they like the freedoms. Changing the things you like to something you don't like so as to win a competition that may come down to little more than a pissing contets seems counter-productive.

In any event, it's moot. The mere fact that open source has the freedoms it has means that choice will simply not go away. Yeah, KDE and/or Gnome may become the "advertising standard" that we use to draw people away from Windows desktops, but unless legislation makes free software illegal, things like X and FVWM and all the other "oh it's so confusing save me from having to choose" things that we hear whining about simply aren't going to go away, because the people who write them want to write them and won't stop in the name of some corporate strategy.

-Rob

Re:A thought

[ClippyHater]

Do we then really want to take over the desktop?

I think having a majority of desktops running Linux would be a huge boon! No more searching to see if ALSA supports your latest/greatest sound card, no more searching to see if the latest/greatest graphics card is supported and has full hardware acceleration in 2D and 3D. When you "own" the desktop, suddenly device manufacturers find it prudent to write the drivers for you.

No more Wine/what-have-you to run some of those fantastic commercial apps under Linux (and spottily at best for some). The manufacturers will find it in their best interest to do a straight port to Linux to get to the most users.

So, IMO, yes, we do want to own the desktop

If we have to become like Microsoft to defeat Microsoft, then what's the point?

If by "become like Microsoft" you mean suddenly having questionable business practices, then obviously you don't. If you mean writing easily breakable software, you don't have to do that either. Why would we have to become like Microsoft??

i disagree

[Tennguin]

This may come as a suprise to many people here but some people LIKE the way linux works. Just because Windows has a lions share of the market doesn't mean it has a superior design... I think a certain company's business practices are more likely the reason why.

I for one beleive that that users would eventually become acclimated to which ever desktop they choose, but that choice shouldn't be stripped away; it part of this communities appeal.

KDE and Gnome act fairly predictably now; I'm not a fan of Redhat's bluecurve at all. Why bother packaging two desktop enviornments at all if both are coded to behave identically?

Ya know what?

[BHearsum]

I don't care if we attack Microsoft's monopoly, or takeover the desktop. Since when do I care one bit about 'the average user'? I'm using Linux because it works for me, if Windows works for someone else then let them use it. If you take away the choice, then to me, you're saying that one size fits all, which is completely untrue.

Besides, there's already distros that have 'standardized' certain desktops for their userbase. Most converts I know are happy with that...

Don't take my desktop away.

Red Hat..

[Chicane-UK]

Would it be fair to say then, that Red Hat has the right idea trying to make a standardised GUI using the bets bits of (predominantly) GNOME and KDE?

Having used Bluecurve'd GNOME over other versions of GNOME, it really is a superb piece of work.. definately the way forward imho, and a huge improvement over the standard.

Re:Oxymoron...

[Ceyan]

Despite the editor's lack of insight into Linux GUIs, that comment was uncalled for. It's comments like those that stop Windows users (be they simply home users, or actual tech savvy people) from switching over to Linux, whether you mean them or not. Grow up.

Third factor...

[Channard]

There's a third factor that should also be taken into consideration - that of just how easy it is to completely mess up an install of the OS. Even if you have an OS that is completely user friendly, making it easy to do whatever you want, if the users have access to essential functions of the system, they *will* mess it up. An ideal OS would be user friendly, secure *and* even the most determined good intentioned meddler would be unable to make a dent in it.

doesn't linux already come with a standard UI

[TonyPyGarthno]

its called the terminal =)

the complaint

[Hitch]

is that there are lots of apps out there that just don't work well without certain UI things installed. I don't have a MAJOR problem with this, but for a while there it was REALLY frustrating to find an app the was KDE only and required the installation of all the KDE libraries and Qt widgets etc. just for a little POS progream. Understand, this was back when I had a 1GB hard drive, and installing all this junk was taking up a lot of room for me - but even now, it just feels like a lot of bloat. Don't get me wrong - I'm not any bigger a fan of Gnome. I don't use either. I do, however, like GTK. as such, I'm far more likely to install the gnome stuff than the KDE stuff. what would be NICE is if gnome and kde were more like "skins" - write a program, include the hooks - and depending on whether someone is using gnome or kde, it comes up as gnome or kde. I know this isn't how these things are written, and this'll never happen, but it'd be nice.

XPDE

[cjcormack]

The Xp.org Desktop Environment is a great idea for people switching from Windows, it's not an exact clone, but will give users a more "friendly" interface (friendly to someone used to windows! not - i've dug myself a hole here... help!!)

My struggle with Linux

[BMonger]

I've installed various distros of Linux (Redhat, Pogo?, and one other (maybe Slackware?)) and maybe it was just my total lack of patience but it seemed like I could get something to work on one distro and not on another. Graphics card would/wouldn't work, ethernet would/wouldn't work, sound would/wouldn't work. I actually started keeping a notebook around to write down the methods I got things to work. Sometimes it'd work again and sometimes it wouldn't.

Then once I got everything working I'd have to figure out which GUI(s) were installed on it. Sometimes they'd work and sometimes they wouldn't. Mostly due to video card issues I'm sure.

Then if I got the GUI to work I couldn't figure out head from tails how to get programs installed. Most everything that I downloaded it felt like I had to build or download from CVS or some weird junk like that.

Eventually I gave up on wasting my time and went back into Windows. Then my Windows machine bombed out (CPU overheated I think) so I scrapped it for parts and now am over joyously running Mac OS X. Yeah it's more expensive, yeah I *used* to have a one button mouse, yeah it looks like a lamp... whatever. I know I have a good and solid OS underneath all those fancy widgets (which is why I wanted to install Linux in the first place) and I have those fancy widgets (which is why I always went back to Windows). Everything works and to get applications installed I just copy them into a directory and voila! Yes on occassion some random freeware/shareware program doesn't work for some reason or another. But overall I think it's a good middle ground between Linux and Windows.

I'm not by any means knocking Linux. I know most a good 25% of the people here probably can get it to run in their sleep and I applaud you for it. But I just don't have the patience I suppose. It's not that I'm afraid of breaking something. It's just that after a weeks worth of trial and error it sorta makes you discouraged.

Wassat?

[fluxrad]

So the argument goes something like this:

"In order to beat Windows, which all Linux users think sucks, they should try to make it more like Windows."

Yeah. That plan's not doomed to failure.

why do we care?

[bokmann]

While I'm all for standards, yes, it is a fact that the goals of the Open Source movement are not the same as the average end user. IF they WERE the same, then I doubt the open source movement ever would have started in the first place.

Why is this a bad thing? Can't we have different goals? While I'd like a little more acceptance, I'm fine with the fact that I will probably always be in a minority of operating system users. I'm also in a minority by having an above-average intelligence.

This is not a zero-sum game... I don't care if Windows or Linux has the larger market share... I just care if I can get my job sone with a minimal amount of hassle.

Best quote...

[blixel]

From the article:

"The average user doesn't know--or care--about the underlying operating system, the idea of GUI interfaces, the various types of file systems, or about any other "technical" aspect of using a computer."

I think this is the best point of the article and the point most often overlooked by technically savvy people. Pick your analogy, driving a car, building a home, operating a microwave or television, etc... The general public cares as little about computers as "we" do about how our cars operate. We just want to get in them and drive.

Blue Curve

[nege]

This is up to the distro owner, in my opinion. There is no standard "Linux" install, you run a distro. And the most popular is probably Redhat, and they DO have a standard look and feel, called blue curve. I think they have done an excellent job with it for the end user, and even though I do not care for it personally, I still have a choice to go download and configure a different WM. Redhat will continue to improve on their standard look and feel, and I always look forward to newer editions to see what they have made better.

Arguments

[GiMP]

1. Linux is not an operating system like MS Windows, it is ONLY a part of the operating system.
2. A product most similiar to MS Windows is a Linux distribution, which IS a full operating system.
3. Users chose their Linux distribution (OS), the Linux distribution (OS) choses their desktop environment.

Ultimately, the user is given a choice of many different operating systems based on linux providing application compatability.

Just because it is different than MS Windows doesn't mean it doesn't work, doesn't make sense, or can't succeed.

Re: every few months...

[Dot.Com.CEO]

Windows has changed aspect exactly once, when Windows 95 was introduced. Even then things did not break - you could stil close a window by double-clicking on its icon on the top left of its window. Almost ten years later you can STILL close windows like that. The changes made in XP are purely cosmetic, a bit more color here, a bit of a curve there - that's all. I have not met one person who looks at an XP desktop after using Windows for years and gets lost. Ever.

Gnome and KDE on the other hand is another story altogether.

Re: every few months...

[Zocalo]

Actually, I think Red Hat really realised the crux of the problem first; it's not the WM that's the problem, it's the look and feel. Hence we got Bluecurve, and now we have a plethora of popular themes being ported between KDE and Gnome. Some of these ports are so good I can log out of KDE, switch to Gnome and my desktop, widgets and applications are identical in both WMs, regardless of whether they were written in GTK or Qt.

Why standardise on a single WM and toolkit when you *can* have your choice and make it?

Exactly

[FreeUser]

How many times have Linus or others said that the goal for Linux is NOT to attack Microsoft's monopoly, but simply to provide a freely usable and stable UNIX-like operating system for anyone who wants it.

Exactly. This entire rant (the Article itself) is basically a characterization of the "People are confused by choice. We should get rid of democracy and diversity, and have one leader, one people, and one empire." Oops, guess I just ran afoul of Godwin.

All of this exemplified the "dumbing down of America" (which is really the dumbing down of the developed world, something Europeans are just starting to wake up to, I think, as this phenomenon is certainly no longer limited to the United States, if it ever was), and the pervasiveness of the mindset that ignorance and laziness should be pandered to, rather than fixed through education, epitomises this.

The point being that, yes, freedom does entail the responsibility and the requirement that you think for yourself. And yes, thinking is WORK. In other words, is Freedom antithetical to laziness? Absolutely. But it is far better to give up the allegiance of the lazy and illiterate than it is to give up our freedom of choice simply to make their lazy lives a little easier.

Of course, the reality is that this false dichotomy is exactly that: false. GNU/Linux neither requires, nor would benefit from, having less choice ("one desktop"), nor does failing to do so make it impossible to appease the lazy and illiterate if that is one's goals (and there are distributions which aim to do exactly that) ... it is sufficient to have one or two defaults (KDE and/or Gnome), which is exactly what we have. I give my friends and family KDE and they are happy with it. I myself generally use KDE, but sometimes I get bored and decide to try out Gnome, Enlightenment, Windowmaker, Flux, or something else. I enjoy that freedom, and I'm not going to give it up (or negate it) just to pander to the illiteracy or laziness of some reluctant ex-windows convert.

A default is enough, and almost every distribution under defaults to one desktop or another. Beyond that, the user can educate themselves and make a choice, or stick with the default, but the idea that those of us who prefer something other than the default (whatever it is ... KDE or Gnome most likely) should give up our projects and devote our energy to working on or testing what others have chosen "on our behalf" is utter and complete nonsense.

1) We aren't out to destroy Microsoft, we're out to enhance our own freedom. Microsoft has become the enemy because they are out to destroy us, and to take away our freedom.
2) It isn't our responsibility to pander to the ignornance or laziness of others. It is their responsiblity to learn, or not, as they see fit.
3) Freedom has built into it responsibility ... it is neither designed for, nor applicable to, those too lazy (or uninformed) to excersize it.

Then again, I always thought changing the verbiage from "Infammible" to "flammible" to appease the ignorance was a profoundly idiotic move...

Re:Exactly

[redtoade]

This post is the epitome of why Linux is failing on the desktop! This argument isn't about Linux! It's about migration.

All of this pseudo-intelligencia, philosophizing garbage demonstrates how isolated most computer science geeks (and thus developers) are from the majority of computer users. Inevitably, the entrenched Linux community balks at standardization. That's just a fact. Somehow they derive their collective self image as software rebels. Linux is the icon for the anti-establishment, "we are not part of the normal world and proud of it," the internet is ours DAMN IT" crowd. Well get over yourselves. As Linux is adopted into the work force, more and more home users will adopt it also. AND THAT'S what this discussion is really about...

This isn't about Linux users! The majority of computer users use Microsoft Windows. So any decent designer would immediately recognize how easy it is to smooth out the learning curve by drawing on the one thing that almost all computer users have in common: familiarity with the gui. Windows users don't care about computers. They don't spend their days dreaming about kernel compilation... and they never will. They typically go to work, do their job and go home when they're done. Most don't turn computers on at home. Most don't have CompSci degrees... or any college degree for that matter. Most just want to use their computer the way they use their TV: turn on, watch, change channel, turn off.

If televisions were packaged like Linux distros: where components arrived in separate shipments, the consumer was responsible for complete assembly, the instruction manual's chapters were scattered all over the internet, with no two television sets having the same controls nor receiving the same channels... there would still be some self righteous jerk screaming how standardization would be the end of the world! And only a small percentage of isolated geeks would be watching TV expecting the rest of the world to lose their technical ignorance. Get real.

The expectation you have of the user to close the distance between what they know and what you want them to learn is naive. That's not how industries produce consumable goods. And like it or not, Linux is now part of that economy. Despite whatever philosophies you chose with which to see the world, supply and demand exists. If you choose to ignore that fact, I guarantee that there's someone else out there that's not so short sighted waiting to take over where you stopped short.

What really pisses me off is the throngs of Linux minions who scream that standardization is limiting freedom of choice. No one is saying that the desktop needs to be locked down to one gui! What is being said is that the default gui should be the same across all distros. That's not very difficult to do despite the KDE vs Gnome debate. You pick KDE. End of story. I'm a Gnome user myself, but let's face it, KDE is more similar to Win32. But so what? I would expect the distros to still have both standard and advanced installation options. The standard would be an out of the box vanilla KDE install. The advanced would give you the choice between desktops. Why is that so difficult? Answer: it's not. It's a pissing contest. No one wants to favor one desktop over the other. Linux users feel their sense of rebellion being flushed down the toilet as more and more non-Linux users cross over. Tough.

You're right and wrong

[aliens]

We certainly can make things customizable/replaceable when starting with a default.

The thing is Linux does does not have a default. And that's his point. I sit someone down in front of Windows they recognize it. The goal is to sit someone down in front of linux and they'd go, "oh it's a linux machine" and know what to do. The problem is I take a dozen linux boxes and each one is vastly different. One runs KDE, one Gnome, one IceWM, etc.

That all said there will never be a default. Redhat will want to do it one way, Mandrake another, Knoppix another. So the entire discussion is moot. Linux will never be the desktop competition for MS. Redhat maybe, or Mandrake possibly.

What you'll hear is not, "oh this is Linux. I know it" Rather it will be "oh this is Redhat"

And that's the way it is.

Lets face the real problem!

[forming]

The whole "Standard GUI" is far from the problem. It may be a small part of the problem, but the lack of a standard GUI isn't what keeps the average user away! I have never heard someone say "I would like to try Linux, but there are just to many choices of GUI's."

I think the one major thing that keeps users away from Linux is the fact that you can't just go to the store and buy some software or hardware and just put it in and it magically works. Sure there has been a lot of progress made in this area but it is no where near what it is for those other operating systems. If more hardware vendors would start releasing drivers for Linux and these software companies would start porting there applications to Linux this would be a whole new ball game. Without some help from the rest of the PC industry Linux never has a chance at cutting in to the MS monopoly.

Developers as well, not just users

[MrEntropy]

I would also add that is is important for developers to see a standard environment as well. For instance, if I am Adobe looking to port Premiere to Linux, which toolkit should I use? Qt/KDE? GTK/Gnome? Which distro should I target with which version of gcc and runtime libs? Red Hat? Suse? You can't just pick one, the user may not have that environment installed, and even if it is all statically linked, it may not behave/look the same as the rest of the user's environment. All of this translates to extra expense and hassle in development, which I suspect is a lot bigger turnoff than any GPL hangups people may have. As much as we may bash Windows for changing the environment, at least the Win32 API has remained consistent for the developer.

We may argue that we have all of the Open Source apps we need, but there is still no decent DV video editor such as Final Cut Pro or Premiere. Photoshop is light years ahead of GIMP in features and usability. Roxio has a very full featured and easy to use CD and DVD burner on Mac and Windows, nothing in Linux really compares. Until we make it easy for the developers AND show a market by attracting home users, I don't think we will see these types of apps ported.

This Article is Rediculous.

[13Echo]

Many open source software programmers aren't writing their programs for the average user. They are writing free alternatives to commercial programs, and generally are writing them because there is no such program that is available on the platform. They are writing them because they feel that there is a need for such a program, because, perhaps, they would like to be able to use such a program. Functionality comes first.

And really... Why should they write them for the average user? The average user has Windows. The average user has MacOS. If the average user wishes to use Linux, they have plenty of options that better cater to their needs. A Mandrake, Lycoris, Suse, or Lindows install will best handle their needs. But, the average user is not going to install icewm (as this goon noted in the article). Why would they?

The writer's comments about "weather or not software will work" are pointless. You don't have to be running Gnome to allow have Opera (QT based) work. You can run Mozilla (GTK w/custom XML) on top of ther QT based KDE. Merely having libraries is all you need. The UI can be whatever the user wants to use. I've not found an exception to this. Sure, things may not look uniform, but that's because QT and GTK (and others) use different libraries for skinning and such. They do, however, work.

As far as I am concerned, the typical "rules" of the "average user" do not apply on free platforms. We write software for our needs. We share it with you as well. Most of us, dispite having some dislikes for Microsoft, don't care if Linux dominates the desktop. If the software doesn't suit your needs, then you're probably using the wrong tool for the job (there are plenty of reasons to use Windows). We want choices though. We won't make your mind up for you, and we don't want to. If that's what you want, then you're using the wrong tool for the job.

Go to Redmond to have someone tell you what you should want.

The reason behind Bluecurve

[yerricde]

If you install the GNOME libs, you can run GNOME apps in KDE with the KDE window manager's window decorations. If you install the KDE libs, you can run KDE apps in GNOME with the GNOME window manager's window decorations. If your distribution has made default widget themes for GTK+ and Qt with similar-looking decorations (Red Hat's default theme is an example), and if GNOME and KDE libs agree to handle drag and drop in a compatible manner, users often won't notice much of a difference between GNOME apps and KDE apps.

"Most" tasks is highly inaccurate...

[cnelzie]

Sure you can do somewhat decent desktop publishing, sure you can handle programming tasks, and quite a few other desktop tasks...

However, you are making the same mistake as everyone that spouts that rhetoric. That mistake is quite simple to overlook, because you likely haven't been exposed to it...

The mistake is the lack of Manufacturing software, like CAD/CAM systems, Quality Analysis systems and other extremely important engineering and design software.

Catia, Unigraphics, Pro-E and other world-class CAD systems simply do not run on Linux. Control software for Coordinate Measuring Machines (CMM) is only available for the Microsoft Windows platform. That software often controls the basic construction of a manufacturing companies IT infrastructure.

It's the idea of 'incompatibility' and the desire to have a homogeneous network structure that 'forces' many companies to utilize an entirely Windows based network.

Get Catia and Unigraphics as well as the other software I mentioned to be fully supported and released on Linux and then there will be nothing stopping Linux from hitting the desks of the manufacturing industry.

Re:"Most" tasks is highly inaccurate...

[iSwitched]

No mod points today, or you'd be +1.

Your point extends far beyond manufacturing packages. The "Gimp is as good as Photoshop" and similar rhetoric shows how far there is yet to go in the realm of apps.

In fact, I'd say that the recent offering from Redhat and others have taken dramatic steps in easing the issues that this article sees as so important. The desktop environment installed by default on these recent distributions is likely to seem very usable indeed to any reasonable person. In my opinion, the consistent GUI has arrived, sure more work needs to be done, but the framework is there.

The problem, as you've put so well, is that people become tied to the apps they use. What they use at work becomes what they want at home. It always seems to surprise the average geeks that rank-and-file users don't want to learn a new, unfamiliar app to do a task they feel they already know how to do, and as one who focuses on usability and GUI design, I say why should they?

Until that singular arrogance on the part of many Linux advocates (even some, dare I say, who read Slashdot) can be done away with, until more people are willing to scream "the emperor has no clothes!" Linux on consumer and (as you point out), many business desktops, is doomed to lag behind.

Standardize the GUI - It's not a bad thing

[Anonymous Coward]

The call for standardization is not anti-innovation. It is about basic commonality of functions and consistent user experience. That's what Apple got right with the Mac early on, and MS caught onto after a while.

I've been a Linux/BSD user on the server side for over 10 years. I love it, but I've never really tried it on the client side. I have too many wild applications that are windows based, and there haven't been alternatives.

Well, this last month, I've tried to use the Linux desktop on the side (Knoppix and RedHat 9). Some of the things that struck me were the lack of Cut & Paste across applications, the random UI menu structures, and non-obvious icons.

While I was pretty impressed with Open Office and the fact that things actually worked (considering my past experience was 8 years ago that's rather expected), it was the utter lack of polish and UI consistency that struck me.

Think of it this way, if the linux desktop is so great, why can't I can copy text from my shell window into my e-mail app and web browser? Why can't I drag a file from the file manager into my e-mail message and have it become an attachment?

Developers: Please work for a standardized functional UI implementation. A common API for cut & paste, consistent top level menu labels, clear and common control icons, and maybe even a structure for drag and drop.

There's lots of textbooks about designing UI's, and sure it's a lot more fun to do your own thing. If you expect any kind of broad recognition for your work, or people to tell you what a great job you've done, then the final product needs to merge into something bigger--a consistent UI across the desktop from the desktop manager to the applications and utilities.

Standardization means I can use your app without having to start from ground zero guessing at what functions mean and can apply knowledge from other apps. Things work the same and can interoperate. These are the things that will make the desktop suceed. These are the things MS learned from Apple and implemented to make Win95 and above a successful desktop (yes, marketshare==success).

Problem is BRANDING, not CHOICE

[alexhmit01]

The problem is the term Linux. And RMS's GNU/Linux DOESN'T improve things at all. It is an issue of branding.

Linux is a kernel. Yeah, the systems use GNU tools for System (UNIX, user space stuff) level things, but that isn't the issue.

The ideal would be to DROP Linux from the branding effort (like MS dropped NT from their branding when they wanted to make it the mainstream system, otherwise consumers don't want to pay for "pro" level software).

If RedHat develeped three platforms:
RedHat Advanced Server (powered by Linux)
RedHat Workstation (powered by Linux)
RedHat OS (powered by Linux) [hey, shell out $20k for branding consultants, I'm not a naming guy)

note: Mandrake, SUSE, and anyone else that wants to play should do the same.

then RedHat would be promoting RedHat as the OS. They could then standardize, and utilize Linux's brandawareness in the "powered by" portion, without this problem.

The "open source" effort is about freedom, and a consequence of that freedom is choice, which I see as a benefit. However, that isn't "useful" for end users.

For Example, you should be able to go into a store, and pick up a CD not for "Linux" (requires glibc X.Y+, Linux kernel 2.X.Y+, etc.), but for RedHat, or for Mandrake, etc., then you would accomplish what this guy wants.

RedHat should have a standard look and feel across their consumer and workstation OSes. You should be able to buy "Redhat compatible" software (requires RedHat 9.0 or higher). Now, tech companies could STILL release "software for Linux," but the box could state RedHat 9.0 or Mandrake 9.0 or higher.

In that case, there is NO need to drop KDE or GNOME or whatever. Hackers can do whatever they want. However, the "commercial" install should include RPMs or whatever for whatever distribution they want.

RedHat should have a logo certification, as should Mandrake and any other players.

To have the RedHat logo, you should have to sport a "blue curve" look and feel.

In addition, the "installation" programs should rethink the options a bit.

Sure, the Server piece should let you super customize it, and maybe a free downloadable ISO "hacker edition" as well. However, installation options should affect applications, NOT libraries.

I should be able to target a certain edition of RedHat (or Mandrake, or SUSE), and KNOW what libraries are installed. It is absolutely REDICUlOUS how many libraries are options.

Real simple, if KDELIB isn't installed in the "base," then KDE apps aren't "supported."

The problem isn't an issue of technology existing (afterall, Windows has had Progman/Explorer replacements forever), it's an issue of the branding.

Getting software for "Linux" requires knowing what libraries are installed.

Getting software for Windows or Mac OS X requires knowing what version is installed.

I may need to have Jaguar or 10.2.3 or higher, or whatever for my Powerbook software. I never need to have a particular optional library installed.

And THAT is why Linux is having trouble on the desktop.

Leave the technologists alone, but "Linux" companies, reorganize your installation/brand awareness if you want the desktop, corporate or otherwise.

Alex

Network effects dominate the computer business

[rsheridan6]

If everyone else is using Windows, you're sort of stuck with it yourself. I'm running a Linux desktop and the problems with it are strictly related to the fact that so many people assume you have Windows - I can't get support from my DSL provider, can't reliably open a .doc, and can't run the CD that comes with a textbook (unless it happens to work with WINE - an iffy proposition).

Also, a few websites don't work (they were tested on IE only). If MS gets much more market share we can expect them to subvert HTML/Javascript with IE only features, which will mean that you have to have IE to access the web. With the demise (finally) of NS Navigator 4, that seems possible to me.

But if we get just 5-10% market share, we cannot be ignored. Only a website run by morons would shut that much of its potential audience out, and people would stop using .doc as a standard, more games would be made for Linux, etc. That's why it's important that a certain number of Joe Blows switch from Windows.

Aqua Human Interface Guidelines.

[cryptochrome]

Apple's been doing this forever, with the latest incarnation being in the Aqua Human Interface Guidelines. Apple has put a TON of money and effort into creating computers that are easy to use. They've tried to promote good interface paradigms and discourage bad ones, both in house and with their developers.

Here Linux developers are confronting one of the prime insights of GUI design - a consistent interface is essential to the user's ability to use different applications. There's no need to rack your brain over learning and remembering every different command in different programs, if they follow a consistent, organized, and intuitive pattern. That's why themes aren't built into the OS or even encouraged. That's why there is countless arguing over how metal-style iApps are bad, and ought to go with the less-eyestrain-inducing pinstripe default. And that's a big part about why Mac OS users are so loyal.

Yet another Pundit who doesn't understand.

[LMCBoy]

Linux is not a product, it is a community.

...the open source community must recognize that its primary goals: freedom of choice, freedom of source code, and freedom to alter applications, are not the goals of the average user.

Who cares? The user can either join our community, or they can stick with their OS product. Yes, it's a shame if they choose the latter, but I want to make this very clear: it's that individual user's loss, not our loss. We gain nothing from users who consider Linux a mere product, we need active community members, not "customers". Why should we kill everything great about our system just to attract some MS customers, who may very well be perfectly happy with their current OS? We are not in competition with MS, even though they may be in competition with us.

Well, that's my opinion anyway, I could be wrong.

Stagnation

[TheAncientHacker]

Let's not forget that (IMO) the "competition" between KDE and Gnome, for example, helps to drive innovation in both; Who the hell wants the Linux Desktop to stagnate like some others?

But in reality, the Linux desktops are stagnant.

The only time any "innovation" occurs is when either Microsoft comes up with changes to their desktop and then, as if by magic, KDE and Gnome "compete" by rushing to see which of them can "innovate" by making an exact duplicate of the Microsoft innovation first.

It's sad that in all these years of Linux, this topic comes up each year or so and the debate always becomes a fight between:

• We should have a single standard desktop that looks exactly like Windows - oh, and Microsoft is evil
• We should all be Free to choose between multiple desktops... that all look exactly like Windows - oh, and Microsoft is evil

Perhaps if somebody would actually offer a BETTER desktop than Windows - you know, one that was easier to use but more flexible - there'd be a reason for actual users to switch to something other than Windows. But that would require innovation rather than blindly following the leader while chanting "I'm an individual" and we know that never happens.

Nobody f***ing Gets It

[xant]

Choice isn't good for the user, it's good for the market. It's true that no user wants to make a choice they don't have to. To paraphrase Marvin Minsky, "The more similar two choices, the harder it is to choice between them, despite the fact that the choice is less important by the same degree." This is indeed the case when presenting the user a choice between Gnome and KDE. But that's not what "choice" really connotes in this case. Gnome and KDE are competing for mindshare, and competition is what makes both of them get better and better.

Each one of them continually tries to one-up the other, to support more and more features that the other is trying to implement. It is the competition between KDE and Gnome far more than the competition between Linux and Windows that drives the goal of finding the Next Big Thing for desktop environments. And both of those environments have introduced features that other desktops did not have, including Windows! Windows XP users: notice how Windows XP puts links to recently used applications in the Start Menu now? KDE has had that for ages. Without the competition between Gnome and KDE, the discovery and implementation of those features would slow down drastically.

As to the ridiculous claim that everyone has to be presented with an interface that's familiar to them, if that were true, Microsoft itself wouldn't revamp the look and feel of Windows with every major revision. Furthermore, if that were true, no invention on the desktop would ever happen! Wildly different approaches (OEone, to name one) must be tried so we can continue to seek the perfect interface, and approaches with minor differences are practically going to be absorbed into the user's mental framework as soon as they're encountered.

Users are willing to learn. They all understand that, when sitting in front of a new environment, they're going to have to learn something new. Some people (in general: younger people) like to learn new technology and welcome new environments as a chance to try new things. Other people resist the idea, but they will still do whatever's necessary to learn to use the tools they have available; that is, whatever's in front of them.

That means that minor differences between Gnome and KDE--and they are minor, when you compare the time to learn them to the lifetime of a typical workstation installation--are irrelevant, and therefore the user's choice between the two environments is irrelevant. Choose for them, it'll work out in the end. Most Linux distros already do this, giving a default which the user can change.

And stop kicking this horse corpse about applications. Every modern Linux distro includes the libraries necessary to run both Gnome and KDE apps, regardless of which environment is on the desktop.

MacOS X has solved the Unix GUI problem

[Get Behind the Mule]

A number of people have already pointed out the flaw in this article: Most people buy their Linux installation from one of the distros, and the distros all install a default desktop -- Gnome or KDE or whatever. As the author himself states, casual users rarely change their interface defaults, so these people will just stay with their default setup. And if they never know or care that they had a choice, then so what.

But I want to add something about the subject of open-source Unix and GUIs. Just a few weeks ago, I finally got my first box with MacOS X. I know, I'm way late saying this, but ... a Unix kernel and a Mac GUI, the perfect computer! Absolute fucking paradise. For the first time in my life, I can work with a sophisticated, well thought-out user interface, and at the same time pop up a bash shell and exploit all of the technical power of a Unix command line.

I'm a big fan of the open-source efforts to build Unix systems, but I must say that they have struggled badly and unsuccessfully at their efforts to create good user interfaces. Sorry, but Gnome and KDE and all the rest really do suck (and don't even get me started on proprietary offerings like CDE). Then along comes Apple, big ol' proprietary closed-source look-n-feel-lawsuit Steve Jobs & Co., and puts them all to shame.

I think this guy with the article may be misunderstanding his own point. The trouble with the GUIs of open source Unix systems is not that there's too much choice. It is, unfortunately, that the open source developers have proven to be very, very bad at building GUIs.

Some of the posters so far have stated quite bluntly that open source developers just don't care about GUIs. So OK, score a point for honesty, I guess. I for one am certainly technically capable enough to deal with the likes of Gnome and KDE, but gawd, why would I want to if I can use something as good as MacOS X? Why make your life any more difficult than necessary? MacOS X is the proof, you see, that a Unix interface doesn't have to be so second-rate; if you try, and you know what you're doing, then you can make the user experience with Unix into something thrilling.

I think there's something else besides lack of developer interest that holds back the user interfaces in open source Unix. There's a whole class of people working a field called usability, who spend all their time figuring out how people work well with computers (and other devices). They've built up a whole field of research, and even run empirical experiments to test ideas, giving test subjects an interface to work with and observing what they do with it.

It seems to me that open source developers and the usability people live in almost wholly separate universes, hardly aware of one another's existence. Apple, on the other hand, has been working with these people for years and it shows. I would like to see Gnome & KDE and the rest succeed, but until they start taking usability issues much more seriously than they do now, corporations like Apple will remain way ahead of them.

Linux != Windows

[Seanasy]

I've always had the impression that, as far as the desktop goes, it's not a question of Linux vs. Windows. I think it's Red Hat vs. Windows or Suse vs. Windows. Linux is just an underlying technology. Nobody talks about Mach vs. Windows -- it's Mac vs. Windows.

Linux is a technology. So are KDE and GNOME. People who will use Linux when it has a mature, suitable-for-the-masses GUI won't need to know or care that Linux runs their computer. And they won't need to know or care if it's the KDE desktop or the GNOME desktop. In fact, such information will probably only confuse them. They'll identify their experience as Red Hat or Suse or Mandrake. Not Linux. Not KDE. Not GNOME. Because they won't care and shouldn't care.

Maybe a Red Hat program will run on Suse. Maybe not. It'll be a lot like OS X compared to other Unices. An average OS X user only uses Aqua apps. But, the more adventurous have the means of running X11 apps and other CLI utiltities. It could be the same with Red Hat compared to other Linuxes. An average Red Hat user might only use apps designed for Red Hat. The geeks will know that they can install extra libs and run all the Suse apps, too.

Standardizing desktops isn't really necessary. One may come out dominant. Natural selection will decide that. What the Linux desktop is waiting for, I think, if for a company to take one of the technologies and brand it. They could make some essential apps a la Apple iApps, drop all reference to Linux, KDE/GNOME and present a single coherent experience to the user.

I think coherent experience is the key without the bother of what libs do I need? That's not the job of the Linux developers or even of KDE/GNOME developers. They're just providing raw materials. A coherent desktop experience should come from someone who can synthesize those technologies and present something new, unified, consistent, unique.

Re: What 'interesting things'?

[n1k0]

> What did you have trouble adjusting?

The panel was most annoying. Can I make my GNOME panel look like this?

In general, GNOME's UI is just barren compared to KDE's. When you view the properties of KDE components, there are generally lots of configurables. In my experience with GNOME, I found the opposite to be true. Walk through KDE's control center, and then through GNOME's. I hope you'll see what I mean.

Another issue was how badly Konqueror kicks the crap out of Mozilla when it comes to speed and responsiveness. I'm not knocking one or the other, I like them both and I realize Mozilla's design as a cross-platform application framework contributes to this. But Mozilla has tangible performance problems, and GNOME integrates Mozilla. This is a problem for me.

> When was the last time you used it?

I used it about a week ago and went back to KDE after a few hours. I was pretty disappointed because, as I said, I really want to like GNOME.

-Nick

choice/freedom as a goal

[bninja_penguin]

I have said it before and will continue to say it, Linux has been ready for a long time for the Desktop, the applications are ready and even a lot of users are ready to use it, but the main goal of Linux is also holding it back: choice.

Okay, this statement, coupled with the main link to the article, which states "freedom of choice, freedom of source, etc. etc. is not what the average user wants..." says to me that Linux is progressing just fine. If the average user does not want freedom of choice, then they should look elsewhere. If my freedom of choice is taken away, then Linux is no different than Microsofts offerings, because Microsoft targets the lowest common factor of computing. By "standardizng" on one desktop, one word processor, or what have you, how is that different than Microsoft? Before you say well download and install the GUI you want, or AbiWord instead of OpenOffice, or whatever, that is already an option for those running Microsoft offerings. At work I have OpenOffice running on Win2K, and not MS Office, but will the average joe do that? I think not. If the average joe wants an average OS, with no choices to make on how it acts, then stick with Microsoft, that is what they excel at. If, however you want freedom of choice, and you do care about how it acts, then by all means use Linux. Linux is NOT about displacing Microsoft from the number one spot for desktop OSes, it IS all about empowering the individual user!
Linus says all the time he's out to make the best OS for his use, NOT for the average joe. If you or the average joe wants to use his stuff, fine, but it's no sweat off his brow if you or the average joe doesn't want to. Linus understands the fact that these machines should be our tools, not the other way around.

We need a king!

[Arandir]

Several thousand years ago, the nation of Israel did not have a king. They were a free people. One description of their political system could have been "anarcho-theocracy". But all the nations around them had despotic kings. So the people went to their wisest member, Samuel, and asked him for a king. He warned them that a king would tax then, draft them, murder the best generals so he could sleep with their wives, bankrupt the economy by building guilded temples, instigate civil wars, and divert public funds for the promotion of foreign religions. But the people were adamant. So they got their king.

Skip forward to the modern day. The people of Linux look around them and see slavery, subjugation, domination and product activation everywhere but in their own OS. They started to grumble. "It's not fair that they live in squalor with MSDN subscriptions to tell them what to do. We need a king as well!" So they went to their wisest and said, "Place a king over us, to tell us what desktop we must use! The people of Windows mock us because we have choice. The people of Apple mock us because we are free. The people of Sun mock us because we don't meekly follow orders. Give us a king!"

And so the wisest relented and gave them a king. And the king sent his soldiers into the land to enforce his will. The new desktop would be Knome. Any who used KDE or GNOME would be banished. Any who used XFCE, Windowmaker or Blackbox would be arrested. Any who programmed for Qt, GTK+, Fox or FLTK would be pilloried. Distributions who offered their users a choice were stripped of the LSB certification.

Window Manager choice?

[JonKatzIsAnIdiot]

"the choice of desktop between KDE, Gnome, IceWM etc, is not one that a former windows user, even a fairly technically competent one, is going to able to make an informed choice on"

Picking your window manager or desktop environment is easy. What so hard about:

1. Picking KDE/Gnome/Enlightenment/IceWM ... from the login menu
2. Clicking on the icon in the lower left corner to see what programs are available (if it's there)
3. Clicking on the icons to see what they do.
4. Trying a few right-clicks to see what happens.
5. Deciding which one you like.

IMHO a far greater problem facing new users is figuring out what software they need. Many free software projects do a terrible job of representing themselves to the user community. When a user visits a program's web site, they want to find out if that software will fit their needs. If the features and capabilities are not spelled out clearly, the user may go elsewhere, or assume that his only option is commercial software.