Ask Slashdot: How Would You Fix the Linux Desktop?

itwbennett writes "Slashdot readers are familiar with the Torvalds/de Icaza slugfest over 'the lack of development in Linux desktop initiatives.' The problem with the Linux desktop boils down to this: We need more applications, and that means making it easier for developers to build them, says Brian Proffitt. 'It's easy to point at solutions like the Linux Standard Base, but that dog won't hunt, possibly because it's not in the commercial vendors' interests to create true cross-distro compatibility. United Linux or a similar consortium probably won't work, for the same reasons,' says Proffitt. So, we put it to the Slashdot community: How would you fix the Linux desktop?"

Comments

It's not broken.

[Hatta]

I've been using Linux on my desktop for 13 years now. It works just fine for me.

Re:It's not broken.

[CodeheadUK]

That's nice, but you're not the target of this question.

It's the learning curve that puts most people off. If you can get the average user through the first few weeks with minimal problems, you'll set them on the path to become a beardy 13 year Linux veteran just like you.

However, most people's experience of Linux is a troublesome couple of days trying to get some obscure bit of hardware working properly followed by a full on feet-eating system meltdown due to excessive fiddling in the wrong places. People (right or wrong) have short attention spans and things need to 'just work' or they'll go elsewhere.

Re:It's not broken.

[Anonymous Coward]

It depends on the attitude. A satisfied user who doesn't acknowledge there may be problems preventing wide-spread adaptation is a road block.

Re:It's not broken.

[Anonymous Coward]

There's no problem. For many a rock is exactly what they need. For many linux is exactly what they need. For many, windows or mac is what they need. Everything has a function.

For me personally, linux is far more functional than Windows in my day to day as a web developer. The only thing I pay yearly licensing for is VMWare so that I can run multiple servers and testing environments.

Linux doesn't need to change to be useful to many people. As people get more bathed in technology from birth, the barrier to entry is going to decrease. We're already seeing that. Distributions like Ubuntu you can almost completely avoid the command line and have an app-store like experience- this lowers the barrier even more. We're there right now. This is the time.

If your concern is foisting Linux on people who are fine with the tools they're using, that's a different problem. You have to overcome in that case. People will come to linux when the price is wrong for other things and when their needs relative to their dollars aren't met.

Don't push linux. you're no better than the assholes that parade around foisting their religion on you. Linux is a tool and it is a religion. It will be found by people who seek it, and every day more and more people are doing that. Linux isn't a foreign term to almost anyone who has an android phone or reads the news. People are less and less afraid of it as they know more and as it looks more like what they know.

Give it time.

Re:It's not broken.

[0123456]

Because most users don't install Windows themselves?

Duh, you've got it, young padawan.

The only reason people think Windows is easy to install compared to Linux is because they don't do it. Take a blank PC and a fresh Windows install CD and see how easy it is to get running.

Re:It's not broken.

[pnot]

With a learning curve like that, why would anyone want to run Windows?

Because most users don't install Windows themselves?

And here we have it: the simple answer.

The way to have more people using the Linux desktop is to HAVE IT PREINSTALLED by vendors, because most people are unwilling to install an OS themselves from scratch, no matter how incredible it is. Of course, this is much easier said than done, but I think that blaming GNOME/KDE/Unity for Linux's 1% market share is missing the point by a mile.

Re:It's not broken.

[Enry]

#%^#%$$ n00bs....I've had a /. account longer than most of you have been using Linux.

20 years this year. I started using Linux on my desktop as my primary OS in 1992.

You know what Linux needs to be 'successful' on the desktop? Stability. Same look-and-feel for the OS across the same distribution over a long period of time. Same set of applications that get installed. Every time I upgrade my OS (and I've done a *lot* of upgrades) the interface changes. Every 6 months I have to install a new OS. Sure, the LTS Ubuntu make it a bit easier, but that just means a larger gap between what I'm running and what is current and what everyone else is using.

But that's the appeal. I mean, I'm the kind of person that wants the latest-and-greatest (not necessarily bleeding edge, but functional). So I grit my teeth, upgrade to Unity, figure it out like I've figured out Motif, Enlightenment, fvwm, Gnome, KDE and every other windowing system/environment and get back to doing work. That works for people who want to use Linux, but doesn't for everyone else. Look at how OS X and Windows have looked over the past 10 years. The look-and-feel is basically the same. There's changes (replacing the start button with a windows logo), but they're nowhere near as drastic or often as you see in Linux. Maybe Windows 8 will change that. Haven't used it yet.

Now what can really be fixed? There's a lot of rough edges that need attention. Bluetooth support is horrible, but doesn't matter so much anymore since everyone has gone with wifi. Ability to view and edit Visio documents, or do real calendaring (I've never gotten my Linux desktop to get a calendar from an Exchange server).

There, done yelling at clouds. Now get off my lawn!

Re:It's not broken.

[cpu6502]

>>>The main roadblock is that the market has been dominated by a single vendor

Then the solution is to copy the vendor. Make the Linux look & feel like the XP/Seven OS that everyone knows and feels comfortable with, so the transition is near-painless.

People don't want to relearn how to use a computer, just as they don't want to relearn how to drive car. (Notice how the hybrids make themselves have the same controls as standard cars, rather than separate controls for the gas engine & electric motor.)

Re:It's not broken.

[betterunixthanunix]

A satisfied user doesn't help "spread the Linux base"? Why not, I ask seriously?

Maybe we do not want to get stuck doing tech support for our entire social circle. It's unfortunate, but it is true: we are still at the point where if we install GNU/Linux on someone's machine, we take on the responsibility of solving their problems (and that ultimately means solving problems that are unrelated to their OS -- an unplugged cable, an overheated router, a power outage, etc.). LUGs are dying and cannot provide useful community help, and online forums are full of bad, contradictory advice. We are still not bothering to educate anyone about computers (except how to use a speciifc company's product in specific ways), and so most computer users remain helpless.

This is not merely an uphill battle; it is more like an attempt to reach escape velocity.

Re:It's not broken.

[Hawke]

(Hey man, long time no see)

This. Like Enry, I've been using linux since pre-1.0. Unlike him, I've lost my desire to constantly upgrade versions.

The "KDE/Gnome are both Windows 95/XP look-alikes" era was probably the top of the usability as far as I can tell. Newer KDE never got back to the same level of usability, and newer gnome makes me turn giant and green. (Look, my monitor is not 1024x768. Stop making UI decisions that only work on tiny-ass monitors.)

And unlike most here, I think that is reasonable. Normal people won't use Linux until the app they want is only available on it... and that won't happen until the developer likes it enough to run it as their default platform. So YES, make it nice for neckbeards first. And once it's (back to being) nice for the neckbeards, THEN go ahead and try and make it nice for your grandmother too... but DO NOT break it for the neckbeards.

And then you declare the basic desktop DONE for 3 years or so, and work on apps. Maintain the desktop in terms of bug fixes, and internal reworks and anything else you need to do, but religiously keep interfaces static for 3-10 years. And instead of going all 2nd system on the interface, work on other things. Maybe those are easier app-building tools? Maybe those are actually just killer apps. Maybe those are better tools for configuring the system, or for managing large numbers of desktops. Maybe that's "work on something completely different that doesn't affect the desktop". Whatever. Maybe that's "work on something completely different, like servers". I don't really care, as long as you stop breaking perfectly working desktops.

Re:It's not broken.

[HaZardman27]

If the solution to Linux's "problem" is to turn it into the crappy OS that it absolutely strives not to be, then I would rather stick with the "problem." I think most of the Linux community would agree with that.

Re:It's not broken.

[chipschap]

Unfortunately if Linux were to look exactly like Windows and work exactly like Windows and lose the multi-media issues and have a lot more apps*, Microsoft would still dominate ... either just through inertia or more likely through a combination of inertia and additional changes. They would change the game and dare Linux to keep up as they leverage their near-monopoly position.

Linux works great for me, a retired long-time computer professional; I'm able to get more work done, faster and better, than I would be able to do on Windows. (And by the way, that "work" today is novel writing. You certainly don't have to have Windows or a Mac to do creative things.)

It also works for my wife, who is an "average" computer user. But then again, I support her system and fix problems (which are about 99.9% to do with multimedia).

*Does Linux really need 'a lot more apps'? Maybe, when we're talking about gaming. But for basic use? With GIMP, Inkscape, LibreOffice, etc., it seems as if the bases are covered. What "killer" apps are required? (Yes, there are industry-specific 'niche' applications that only run on Windows; I'm talking more basic than that--- and a surprising number of the 'niche' apps have Linux near-equivalents.)

Re:It's not broken.

[aussersterne]

"a troublesome couple of days trying to get some obscure bit of hardware working properly followed by a full on feet-eating system meltdown due to excessive fiddling in the wrong places"

That is not a learning curve. That is refusing to separate the role of developer from the role of user, which is the primary characteristic of the Linux community.

This comes up every time there's a story about security on Slashdot ("they shouldn't be allowed on the net without first learning...")
It comes up every time there's a story about a Linux project ("...don't like it, you can write your own...beauty of open source...what have you coded...")
It comes up in every story on GNOME or KDE ("...fixed by extensions...prefer choices to no choice...")

Blah, blah, blah.

Users are not developers. Every product that wants to be successful amongst users must treat them as users. Users want:

1) Full functionality out of the box.
2) To apply tools toward other problems (not to apply their own labor toward tool maintenance/creation).
3) A sensible basic tool configuration/set of properties that never needs to be changed.
4) Respect for what they're trying to accomplish.

Linux provides none of these, 20 years on. From the user's perspective, it is thus broken.

- In many cases it doesn't work out of the box.
- In most cases *some aspect of the system* doesn't work out of the box.
- Their requests for help are met with instructions to apply themselves toward learning more about how the tool is/was made and toward improving the tool itself.
- The defaults are almost always wacky. No distro or desktop has really ever shipped with good (non-ideological/non-developer) defaults to this day.
- Users are constantly condescended to, as though anyone whose primary task isn't Linux software debugging/development is a worthless n00b.

Here's how to fix the Linux desktop:

- Stop focusing on OS development pie-in-the-sky and call the core OS and desktop implementations and APIs good enough. Stabilize them for a decade at a time in this "good enough" state and allow bugs to become "known issues with workarounds" that can be used for a decade at a time.
- Pour development hours into consumer-level/user-level stuff: multimedia, graphics and audio support, broad-based hardware and driver fixes.
- Stop "shipping early and often." Ship late (i.e. once bugs have been fixed/stabilized) and rarely (no more than once every couple of years).
- Stop providing "learning curve" instructions. If they have to resort to dotfile edits or man/info pages, just say "Linux can't do that yet for users" instead. (Yes, it can do that for developers, but developers are not users.)
- Stop the "free software" puritanism. If something that's needed can be licensed and included on a "free as in beer" binary basis, and it can't practicably solved with OSS software in time for ship date, include the "free as in beer" version. This goes double for vendor-supplied hardware drivers.
- Create a desktop kernel fork. Linus & co. are not in the business of writing/maintaining a desktop kernel. Their goals are larger (and smaller) than that. The desktop kernel can track the mainline kernel, but shouldn't adopt every latest ABI or other change—just do a major update every 3-5 years.
- Value polish. Stop making fun of "flashy" and "shiny." Consumers buy shiny things. I buy shiny things. People here may prefer a rusted out pickup truck with a working winch to a shiny new performance sedan, but the market for rusted out pickup trucks is relatively small. People want a clean, neat, orderly world, and their computing world is a part of that. The non-developer that keeps clean windows and clean carpets wants a clean and beautiful desktop visible in their living room (and living in their consciousness), not a cluttered black console screen or rainbow-technicolor KDE icon sets with twelve different sets of widgets for twelve different apps. Visuals matter to people and are part of the larger c

Re:It's not broken.

[Grishnakh]

Make the Linux look & feel like the XP/Seven OS that everyone knows and feels comfortable with, so the transition is near-painless.

The problem is, this has already been done! KDE4 works very much like Windows Vista/7 with some minor differences, and is highly configurable and themable to make it look like a near-clone if you want. However, the Linux distros don't like KDE, and are either pushing Gnome3 or in Ubuntu's case, Unity, which are both radical departures from the XP/Vista/7 type interface that Windows users are all comfortable with. The distros seem to think they need to push something new and different and "bold", and that somehow this is going to make millions of Windows users dump Windows and switch to Linux, rather than providing an environment that's an easy transition.

Re:It's not broken.

[Grishnakh]

Yep, I remember having to do a work project with Mac OS 9, and that also had MANUAL MEMORY MANAGEMENT. It took me a while to figure out why my Perl program wasn't working right, until I found out that I needed to increase the memory allocated to the interpreter. Huh? Since when do you need to tell an OS how much memory a program is allowed to use? I don't think even Windows 3.0 had this limitation.

Re:It's not broken.

[rastos1]

I've been using Linux on my desktop for 13 years now. It works just fine for me.

You're part of the problem.

If you want to help spread the Linux base, such an attitude doesn't help.

Me: I don't have a drinking problem.
You: It's worse than I thought. You are in denial!
Me: ???
The truth is, that I'm also a happy Linux desktop user for over a decade. And the only thing that it does not do for me is to compile code using Win32 API. Meh.

Re:It's not broken.

[AlXtreme]

The problem is actually bugs in the handling of tables and form data from .doc/.docx files in LibreOffice.

This. A thousand times this.

Businesses and people are used to MS Word. They don't like it, they simply cope and can be fairly sure a document created in it can be read and changed by another user without too many problems. It's the defacto standard, and any alternative needs to deal with that standard.

In spite of all the effort gone into Star/Open/LibreOffice compatibility over the past 15 years it has always been hit-or-miss when opening a Word document. As long as you can't rely on a compatible Office suite users will simply stay with MS Office, and thus Windows/OS X.

The average user uses their PC to browse the web, read their email and as a fancy typewriter. Get that working properly and you can play ball.

Re:It's not broken.

[jgrahn]

I'm a Mac user. Linux desktops are bloody intolerable for me, because they mimic Window's crappy UI badly.

There are many Linux desktops, and most don't look like Windows at all. Twm-style desktops *predate* Windows 3.11. Tabbed window managers like xmonad are what all the young people at my workplace use, and they don't look like anything else I've seen.

I have to get work done

Part of my problem understanding the problem is I don't understand the relationship between "desktop" and "getting things done". I need

• a window manager (preferably with virtual desktops)
• a way to launch my favorite applications
• a way to shut down the computer
• a way to use my photos as the desktop background

Everything from a classic 1989 X11 desktop to Windows Vista does that. What else do you need to get work done, in the responsibilily area of the "desktop"?

Audio

[Issarlk]

Fix the damn audio and stop shoving a new sound daemon/system down our throat every year.

What, exactly, is broken?

[pstorry]

The problem here is the assumption that something is broken.

Generally, the Linux desktop is fine. There is a choice of UIs, sure - and recent developments in KDE then Gnome haven't helped much. Big changes made people say it was broken - but over time, it seems to settle down.

And with the competition (Apple and Microsoft) also making changes to their desktops, Linux is hardly unique here. We seem to be in a time of change, where people have been challenging the old paradigms. Apple are being the most conservative, Microsoft the most radical, Linux is somewhere in between.

Hardware support? Not necessarily a desktop job, but I'll address is anyway. Linux can't do jack here without more support from manufacturers. When I installed Windows 7 on a (then) new Sandy Bridge motherboard, it found NOTHING. It literally booted into a low res desktop with no sound or network. Only the large collection of driver CDs saved the machine - Windows had nothing to do with it.
Support of Windows from the manufacturers was the key factor.

So let's not bitch about Linux's support of hardware - let's get it right, and bitch about hardware manufacturer's support of Linux.

Apps? We've got plenty, and are getting more. Some commercial apps (Corel Aftershot Pro, Sublime Text 2, VMware are ones I personally use) support Linux as well as Mac/Windows. It gets better every month, when it used to get better every year.

And I guess that's my key message. "You've never had it so good". You may not feel that way, but Linux is on a roll right now, and the question is not whether or not it becomes a 'usable second option'. It's already usable.

The question is whether or not it becomes a SUPPORTED second option - by OEMs, hardware manufacturers, and software companies.

And the signs are getting more positive as time goes on.

Re:Add Support for Visual Studio

[MrEricSir]

Eclipse and others (Anjuta, KDevelop, Kommodo, emacs, etc.) do just fine.

That's frankly the biggest load of crap I've heard all day. You're comparing a professional development tools to Anjuta and KDevelop? For fuck's sake.

The attitude that these half-baked, ancient development tools are as slick as what MS and Apple are offering sums up the problem with the Linux desktop: a steadfast refusal to stay competitive and serious delusion about why the Linux desktop hasn't caught on.

This is asking the wrong crowd to comment

[Hollywud]

You need to be asking this on some other board. Slashdot users are power users and thus cannot bring themselves to get into the 'everyday' user experience. I know they use systems other than Linux, but it's the mindset that is different. I teach this very thing at a University and it is extremely difficult for developers to get into the 'user' experience. That's not a bad thing, it's just a different animal - most users don't understand the things that most Slashdot users will take as common knowledge. If you really want to know then take a survey on a more general site.

Re:It's not broken. Do installfests!

[quixote9]

Preinstallation, preinstallation, preinstallation. That's all that matters. Preinstallation with icons already on the desktop. Why do you think Microsoft fought so hard and long to keep anybody else's browser icons off their precious desktop? Why is the stupid desktop icon worth any price to companies who want their commercial crapware pre-installed?

People will use whatever is in front of their faces. Linux is never in front of their faces. It's not commercial, there are no kickbacks, so it's never going to be in front of their faces. Business IT departments want an 800 number they can call and scream at when things go wrong. Linux has no 800 number. Business IT depts aren't going to demand it, no matter how much sense it makes for the business.

So is it all hopeless? I don't think so. The only thing we can do, you and me, is hold installfests. Help people over that initial hurdle. I've gotten about ten people moved over to Linux (ubuntu) in the last four-five years purely by doing installations for them. And they're thrilled. No more virus problems. Everything works. They're not worrying about the artwork or whether it's a "modern" interface. If we could propagate the get one - install one meme, you can calculate how long it would take for every desktop and laptop to run linux.

Re:Mod parent up.

[Atzanteol]

This is the thing. There *are* times when choice is a bad thing. I think we'd be much better off with one *bad* sound system than 4 competing ones. Seriously - I'm a linux geek and I have trouble just getting sound to work sometimes (and having multiplexing). WTF?

Listen to actual Linux Desktop users

[TopSpin]

Stop alienating power users. We're not the problem. We're the beginning.

If I and hundreds or thousands of others tell you that your desktop doesn't provide the configuration capabilities we need then listen and provide the configurability we're asking for. If we tell you your crazy bloated akonadi/nepomuck/whatevertheflip is too big (a mysql instance in my home directory??) then listen and rethink your design. When we complain that your latest major release is a fabulously buggy mess (KDE 4.0) then listen and don't do that to us again. When you hear from people that want a regular orthodox file manager then listen, provide one and don't deprecate it in favor of some granny-safe photo album browser.

It's not hard, really. It just isn't a lot of fun. Which is why it doesn't happen.

Re:Add Support for Visual Studio

[ezakimak]

I agree. What additional apps are lacking?

- huge behemoth office suite that interoperates w/the defacto standard? libre office. check.
- popular, familiar browser? firefox, check.
- cross platform gui toolkits? QT, others, check.
- *stable* API? I dunno what people are complaining about. The standard c library and POSIX OS API have been stable for ages. check.

I think the bigger problem is *too many* apps included by default--no defacto standard across distros. (But this is what *choice* brings us.)

Users cannot sit down at any linux machine and expect the same experience. They can't expect to always find:
- IE
- office
- outlook
- msn
- notepad
- ms paint
- solitaire (seriously--it's one of the most commonly used apps in the world)

in the same place and working the same way.
There's no IE browser, but there could be konqueror, firefox, chrome, opera.
There's no MS Office, but they might find kofifce, libre office, abiword, etc.
There's a dozen possible IM clients.
There's a dozen possible text editors.
There's no IE GUI file explorer--but there could be konqueror, nautilus, dolphin, or other
There's a half dozen paint programs that might be there.
There's no outlook, but there could be kmail, thunderbird, or a few others.

And that's just the common apps.
They want to install something else, they immediately find:
- there is *no* consistency between distros
- the app they want is probably not ported to Linux in the first place (guess that answers my question--"what apps?"--well we just don't know, but can't expect every dev to make every app cross platform for our favorite platform)

This doesn't even consider what developers have to do to target different distros. RPM, dpk. portage, etc.

Linux is certainly for the most part *source* compatible with a stable c library and POSIX API. But any number of combinations of library versions could be found on a target system--which is why any time I've ever got a commercial app, it usually came w/static linkage so that it would just work.

I see no difference between linux distros and the developers behind them--they are all cats and you cannot herd them.
It's a problem, that by the nature of it's participants cannot be solved.
Desktop Linux will largely remain for developers, by developers--or for people closely related to developers/admins that will install and maintain it for them, or for tinkerers. But not average-day Joe and Susie--it's not consistent enough.

Users no longer anticipate sitting down to a computer system and having to learn it/figure it out.
They expect uniformity to the commodity systems in existence.
*Unless*, they *know* it's some new system and are expecting to figure it out--but in that case, they *expect* it to be the same everywhere they go.
Thus "Linux" is not the best name to use. Really, you'd have to distinguish by saying the distro name. Eg, "Android"--everyone knows what to expect--there are small variations, but they all work essentially the same--no worse than differences in version of Windows.

Thus, "Linux on the Desktop" is a misnomer. It really should be "Ubuntu on the desktop" or "Suse on the desktop" or "Debian on the desktop". So really, it's two problems:
1) which distro is defacto standard (there will never be one--intractable problem)
2) for said distro, what keeps it from becoming a good desktop alternative to Windows and Mac? (see problem #1, and issues above)

QED.

[aussersterne]

Between your post and mine, the dichotomy/disagreement has been made clear.

There are two views of users, computing, what computing is for, and what useful computing actually is at work in this discussion. Another way to say what I was saying is that broader Linux community's ideas of what computing is for and what a user is like are very different from the ideas that are in the economic mainstream.

Rather than respond to your points, I'd like to draw them into relief and point to them. You've made good points with respect to a particular set of goals and a particular value system. But the continuous questions about Linux on the desktop that we see on Slashdot suggest that there is some ambivalence in the Linux world about the ways in which meeting these goals and these values does not seem to lead to widespread adoption.

The stalemate (a decade-old, at least, one) is crystallized by the way in which the Linux community does not want to change its goals and values, yet wants somehow to enjoy widespread adoption. The two are not compatible; to enjoy widespread adoption, Linux must share the goals of the people walking around Best Buy right now. If the broader community wants to distance themselves from these people and these goals, it is destined to fight windmills for a long time when it comes to widespread adoption.

Better, to my eye at least, to simply concede on that point and enjoy the system that exists, understanding that for the limited userbase that it has, it is probably currently the best choice.

Or: You can have users that are not developers or you can have users that are also developers, but there is a distinct limit on the degree to which you can have both groups with the same product.