Asa Dotzler on Why Linux Isn't Ready for the Desktop

An anonymous reader writes "Asa Dotzler of The Mozilla Foundation compares the explosive growth of Firefox to the anything but explosive growth of Linux and what it needs to do to get there for the "regular user" AKA mom, dad and grandma Bootsie."

Comments

Well, here's my take

[the_mad_poster]

My general take on Linux, take it or leave it or try to convince me why I should change my outlook.

Linux is not a bad system, it just doesn't have anything to offer that its competitors don't already do as well or better.

The problem with Linux is not that it's not production ready, it's that it's a system that doesn't have anything special to offer and has nowhere new left to go. It has taken a large chunk of the market share away from the old, cumbersome UNIX systems, with their painful licensing models and lackluster support, but now it has no more market share to chip at because the supermajority of disk space that is left is in the form of desktops.

And Linux is just nothing special in that realm.

I speak authoritatively on the subject because my experience with Linux begins many moons ago with an old system called Linux Mandrake, now called Mandriva Linux. It started with version 5.2, a system forked from the Red Hat 5.2 release. I have since used Mandrake 6.0, Red Hat 7.0 and 7.3, 8.0, 9.0, Fedora Core 2, and variations from SuSE.

The first version I used was painful. It was a horrible system with a horrible interface and horrible documentation. Managing it was excruciating, and it wasn't uncommon for a seemingly simple change to break numerous systems in unrelated modules and drivers. The GUI was weak, disorganized, and difficult to manipulate. The desktop was hard to customize, and the interfaces were slow and cumbersome. Installing and uninstalling was nearly impossible because packages scattered files across a confusing, oblique filesystem, and it was a very common occurrence to find rpm entries had been corrupted and left unusable.

These problems I experienced were not uncommon and plagued Linux for years, leaving astute IT professionals shaking their heads, and young, energetic, and idealistic kids suffering under a burdensome system. I think it is fair to say that the rise in Linux use during the IT bubble and the subsequent pop of that bubble is not a completely coincidental correlation. Literally millions of man hours were lost in this time to troublesome Linux boxes and that sort of loss can hit new IPOs hard when it comes time to pay the piper.

It took many, many years and thousands of developers, but the system finally began to shed its inadequacies and "quirks" and develop into a full-fledged corporate workhorse. The managers who had been shaking their heads warily approached new versions and their confidence was bolstered as the GUIs began to fill out, the quirks began to shrink to the background, and more application support became the norm on new releases.

Now, Linux is a force to be reckoned with in backoffices and server racks. It is not, however, any closer to dethroning Windows as the supreme ruler of meatspace userland.

There is a very simple reason for this: it sucks.

I know, I know, I just finished zipping up the body bag on the "Linux isn't production ready" myth, but we've moved to a whole new realm here. We've gone from the terminology of fsck to frag. From SMP to MMORPG.

The problem is that everyone knows Windows and everyone's applications already run on Windows. There is no purpose in learning a new system because Windows is now polished and stable, and maintains its original attractiveness through its continued ease-of-use. Like Linux, it has shed its inadequacies and become a competent and powerful system in its own right.

So, in effect, we have the Windows system which has provided a consistent and simple interface for a decade now, and the Linux system which is an alien world to most people. Both function competently, though Linux still suffers a bit from the problem of glut thanks to its monolothic structure, and neither really offers a serious bnenefit over the other. As Joe Sixpack sitting in my cubicle, I have to think "Well, then why should I switch?" As the IT manager evaluating the cost of switching, I have to ask, "Well, how can you justify the tens of thousands I'll need to spe

Re:Well, here's my take

[wolffman1982]

I don't think the collapse of the irrational speculative bubble of the late 90's had any real correlation with those start-ups using cumbersome Linux systems. Rather, the reason peapod/pets/...com failed was because they were given amazing amounts of money to fund business models that didn't make money.

The VC firms just wanted a company to grow as fast as possible in order to create a profitable IPO. Linux costs were a pittance compared to the misappropriation of money by these .com's.

Re:Well, here's my take

[KaptNKrunchy]

Plus, Jews use Linux.

Duh, it's free! (my appologies to my fellow jews, just couldn't resist)

Re:Well, here's my take

[aussersterne]

Classic formulation: if you're not interested in adopting the Unix mindset (text-based text processing, pipes, small well-defined tools, a de-emphasis on graphical user interfaces, non-data-processing devices, etc.) then why choose a Unix operating system?

Linux offers a great deal of value that Windows doesn't. As someone who works with huge databases of text at a major publisher on a day-to-day basis and who has to use both systems at varying times, I can assure you of this. Just because you don't have the needs that justify the Linux learning curve doesn't mean that no-one else does. And even if you can't even see any features that Linux/Unix has that Windows doesn't, it's fairly rich of you to assume that everyone who chooses Linux/Unix over Windows does so simply becuase they are deluded.

I can honestly tell you that for any number of large jobs in my workplace, two or three commands at a Linux command line replace either dozens of labor hours, dozens of development hours, or the $$$ to purchase a specialized product in Windows.

What I don't understand is why desktop users who have no need of the "Unix philosophy" of data processing insist on complaining about an operating system that was designed to move DATA (not icons or mouse pointers) around efficiently.

If it doesn't fill your needs, don't use it. The unfathomable leap comes when you assert that no-one else should either.

Re:Well, here's my take

[Eric Damron]

"Linux is not a bad system, it just doesn't have anything to offer that its competitors don't already do as well or better."

It's not what Linux has that attracts me. It's what it has not.

It doesn't have a philosophy of lock in.
It doesn't have virus after virus taking my system down if I don't patch it daily.
It doesn't have a philosophy of limiting my fair use rights.
But most of all it doesn't have Microsoft with their anti competitive practices and their obnoxious licensing agreements.

Re:Well, here's my take

[Randseed]

Pretty sure there is at least 1 windows box running in a local hospital. Police stations don't use computers, no? Well, I guess these aren't important after all.

I might as well comment on this since health care came into it.

Most hospitals use Windows exclusively for user interface boxes. Their backend systems use things like AIX, Linux, or some other flavor of UNIX to run the databases and such. The catch is that universally these user (Windows) systems are overloaded, cumbersome, full of security holes, and a total mess.

Take a real life example of a hospital system that has one major inpatient facility, and a bunch of outlying clinics. Their backend uses some UNIX variant, and they use some Windows program to interface with it. So far not bad. Enter the problems.

First off, there is no unified user home directory system. (Yes, Windows can do this.) Under UNIX, the simplest way to handle it is do a networked stock setup of all the user machines, mount home directories under NFS (or preferably something better), and avoid a lot of problem. This helps with data security, because patient information (e.g., the progress note I just typed in Word) is saved under my user account with default permissions of -rw-------. Second, this means that if I go to one of the outlying clinics, I can still access my files.

I've actually thought really hard for a few years now about why these hospitals blow literally hundreds of thousands of dollars on Windows machines, when they could set up Linux machines that are better administatable for much less. After they set this up, they have to deal with all the random crap of people violating the security protocol by installing God only knows what (Gator, anyone?) which, even if it affects only their account, is still a problem.

But perhaps the biggest problem with Windows in these environments is this: The things take so god awful long to log a user in (two minutes at least at my last place of work on a 2GHz box) for whatever reason that much of the time people don't even bother to log in and out. That leaves a terminal sitting there, and makes the password protection at the OS level worthless. (Yes, obviously something is wrong with all these installations -- probably Novell -- but it's universal. I've never seen a Linux installation that bad, EVER, in ten years.)

Re:Well, here's my take

[drgonzo59]

Same goes for some of the medical equipment. Some MRIs and such have often some UNIX variant installed to control them. As a general observation I would say that in a critical environment Unix/Linux systems are preffered over MS Windows. Somehow just that fact says something about the qualities of both systems.

I used to have a Windows partition until 2 days ago when my IBM 60GXP "Deathstar" died with the infamous click-o-death. I don't think I will bother installing Windows again. The _only_ thing I was using it is to play the latest games, so I'll have to do without that. My main operating system is Ubuntu Linux - very nice distro, fell in love with it at the first sight. For me Linux reached the point where I would use it on a Desktop instead of Windows.

Re:Well, here's my take

[It'sYerMam]

I think at least part of the problem with all of these "Linux isn't ready" arguments is that there is so much focus on getting "Joe Average" to switch over. I truly wonder whether Linux will ever be the right option for Joe Average - there's no doubt that with a system as customisable as Linux, there will be added complexity, and to reduce one is to reduce the other, at least at the surface level. The challenge for people like the GNOME guys has always been to balance user-friendliness against customisation.

However, would it not be better to ignore "Joe Average" or "Joe Beloe Average" at any rate, and focus on people who are better able to handle the technicalities? Of course, that doesn't mean a regression, nor ignoring Linux's current failings (such as dependencies) but it does mean moving the focus away from getting everyone and his grandmother to migrate.

Linux Objectives

[fembots]

It seems Linux has to be "like Windows" to attract a broader user base. I thought that is obvious and has been talked about for many years.

The thing is, how many of the developers are willing to sacrify what they have built so far in exchange for a bigger market share? Are linux developers really keen to get as many people onboard at all cost?

I guess what I'm trying to understand is, what are the objectives in Linux? What is it trying to achieve? Is world domination still the name of the game?

Re:Linux Objectives

[bedroll]

Linux doesn't have to be any more like Windows than it has to be like Mac OSX. If devs try to make consistant and logical interfaces for their applications, and make those applications useful, then people will use them. I don't think developers should change things to be more like Windows, because they should be trying to be better than Windows.

My opinion on home desktops: Microsoft can keep them. I don't want Linux to be screwed up by attempts to make it work like Windows. I don't think Linux needs to have any more of a userbase than it already has. If it gets it, then great. If it continues on the path it's been taking then it'll probably have a broader userbase anyway. Why should the FL/OSS community worry about rapid adoption? We know that we'll be around later. We just need a large enough userbase that it can't be ignored, we need it for interoperability and nothing more.

I don't agree with everything that the article says. I don't think that the desktop experience for a GNome or KDE user is so perplexing that they can't figure it out. More likely they'll have no problem at all with anything if whatever administrator that installs the OS shows them where a few apps are.

This all comes back to the user being separate from the administrator. I don't want to use an OS that confuses those two (like most Windows setups do). I'm happier using an OS where someone with a clue is the administrator and the user is kept out of those tasks.

Re:Linux Objectives

[Mornelithe]

Was the "objective" of Linux ever world domination?

I thought it was a bunch of people working in their spare time (and more recently, for corporations) to produce a Free operating system that they would enjoy using/developing, and, in many cases, that other people would enjoy using as well.

Many people have some political agenda when they work on Free/Open Source Software. Many people don't. Asking "what is [Linux] trying to achieve," isn't a very well-defined question, because "Linux" is not a single sentient entity. It's a community of people with many different goals and ideas.

But they do...

[HappyHead]

But my mother and father already use Linux - mostly for the games though.

Fundamental assumption?

[aussersterne]

All presupposing (as is so often done) that the ultimate "goal of Linux" (thereby attributing to "Linux" intentionality that it no doubt does not possess) is to woo Windows users away from their desktops, rather than to provide a superior computing and data processing platform.

I am very happy with the latter, which Linux has provided me with for some years now, and if Linux ceases to do so in favor of attempting the former, I'll happily switch to some other platform (until "I hate elitsts" n00bs who want to be elite but don't want to work for it invade and begin to transform-to-inefficiency that one as well, at which point I'll move on yet again).

Give me efficient computing or give me death. I want to manage my reams of data and my network tasks. I don't care if it jives with the [utterly inefficient] way of doing things in Windows, or if the Windows users care to adopt my methods.

I just want the powerful tools, unpolluted, task-oriented, intelligently designed, that let me talk to my computer using the language through which it can most quickly and subtly be isntructed.

It's not an elitist view, it's the view of a data processing pragmatist with a lot of tasks to juggle and a lot of work to get done.

When you get right down to it ...

[ScrewMaster]

Ma, Pa and Aunt Bootsie are irrelevant. The corporate world is where the money is, and that's the area where Microsoft is most concerned about losing market share to anything or anyone. Right now, people that buy a computer for home use are to a large degree constrained by what they use at work, which is most likely Windows. All this talk about Linux being ready for Joe Sixpack belies the fact that operating system acceptance begins in the workplace and filters down from there. If the idea really is to displace Microsoft, then the place to start is the cubicle farm, not the den. The original IBM PC, all those years ago, gained widespread popularity among the corporate set because it had a ready-to-go set of business applications (and, of course, the IBM name.) Everything else flowed from there ... and it's still true today. Linux really needs (and is getting) some heavy-duty office/business applications and functionality. That's what it will take.

Don't confuse the market segments.

[khasim]

There are 4 major market segments:

1. Servers
2. Corporate/government desktops
3. Mom/Grandma home users
4. Power users/Gamers

Linux is making huge gains in the server market. The statistics show that.

Linux is just starting to gain in the corporate/government desktop market. Expect this to take at least another 3 years.

Once OEMs are comfortable with Linux (due to large orders from corporations/governments), they will start offering it on desktops suitable for basic email/web surfing. The largest limitation is lack of drivers for new hardware. As this market grows (slowly), that will change.

Which will, finally, result in power users and gamers having Linux as an option. That means that the latest hardware will be released with good Linux drivers and the games will be available on Linux. The biggest problem here is the Microsoft desktop monopoly.

Other than that, a corporate KDE or GNOME desktop can be made to look almost exactly like a Win2K desktop so there is no need to worry about training the end users.

The value of Linux doesn't exist for the last two market segments (both home segments). The value exists for the server market and the corporate/government desktop market. But that value will drive the home adoption as people become familiar with Linux at work.

The original article is correct in that having a way to capture the info from Windows would be a major boost to Linux adoption in the home segments. But without the hardware/game support, it just isn't worth the trouble for the average user.

Firefox is worth the trouble of the few websites that don't support it because of all the great features of Firefox (no ad/spyware, very few popups, ad-blocker, etc).

Re:Don't confuse the market segments.

[Synbiosis]

Which will, finally, result in power users and gamers having Linux as an option. That means that the latest hardware will be released with good Linux drivers and the games will be available on Linux. The biggest problem here is the Microsoft desktop monopoly.

Uh, no. I'm one such power user, and the problem with Linux is that moderate customization requires intimate knowledge of the command line and Linux's quirks.

I'm an XP man, but when I used a PowerBook for two weeks, I could easily install and remove programs, connect my external hard drive, and had some interface customizations up and running in a matter of minutes.

Compare that with my Linux experience: Two months ago, I installed Kubuntu onto my laptop. It's very likely that all of the issues that I ran into are easily fixable, but the solutions were simply not apparent or mentioned on all the help sites I went to. Let me tell you one thing: Most people don't like recompiling the kernel, compiling programs, or compiling drivers. It's probably a simple process (I've never had success with it myself), but it simply should not be required for usability purposes.

First issue: Installing software. This blew ass. First, I had to find out the name of the package, and tried to use apt-get. This didn't pan out. Then I found out that the servers that apt-get was trying to use were disabled, or something to that effect, so I had to go and edit a text file to allow for this. This pissed me off quite a bit, because had I not been dual-booting Windows at the time, I would have had no way of knowing how to fix it.

Second Issue: Getting my wifi card to work. This was fairly important, as my connection utilized my campus's wireless. So whenever I had an issue, I had to reboot into Windows and search for it. I never managed to get it to work, even though I have one that's fairly ubiquitous (Intel 2100). After futzing around with the command line for a couple of hours and browsing some sites, I tried to figure out how to install the drivers derived from the intel open-source release. Then I foudn I'd have to compile the drivers or whatnot, and I gave up there until I could find someone experienced in the matter.

Third Issue: I couldn't get it to sleep. I spent a good 40 or 50 minutes to find out that I needed to recompile the kernel to include support for sleep.

Fourth Issue: Realizing that I had just wasted my time installing Linux. I could do everything I could do in Windows, except it took me twice as long. Screen space was a serious issue. Using OpenOffice at 1024x768 felt like using MS Word at 640x480. My screen always felt cramped. Image editing in GIMP just sucked. Even when using Photoshop, I felt that 1024x768 wasn't sufficient for some of the stuff I was working on, and using GIMP made me feel like I was working at 800x600.

Obviously, I'm back to XP. I'm willing to spend the extra $300 to save the time and frustration that I've had using Linux. I realize that pretty much all of these issues are 'trivial', but the fact that I ran into all of these problems in the span of two days seems unacceptable. When I used OSX for the very first time, it took me less than two hours to become proficient. XP took me 4. Linux needs to get to the point where a power user like myself can be quite proficient over the span of a day.

Re:Don't confuse the market segments.

[tehdaemon]

Have you ever tried to run Windows on a laptop without the drivers and utilities provided by the laptop maker? I have. The drivers were for XP home and we were trying 2K. Almost nothing worked. We then tried Fedora core 4 on the same laptop. While there were a few problems, most stuff worked. I am posting from Ubuntu, it is easier than fedora (IMHO)

The problem is not linux, it is manufacturer support. Windows is actually much worse than most linux distro's, but because the manufacturer supports the laptop, it (usuallly!) works fairly well. If they gave the same support to linux it would work just fine.

Re:Don't confuse the market segments.

[tokabola]

Uh, no. I'm one such power user, and the problem with Linux is that moderate customization requires intimate knowledge of the command line and Linux's quirks.

That's funny, I set up a half dozen systems with Mandrake 10.0 and never had to use the command line once. Sure, there are some things that can't be customized without using the cli, but making those same customizations in Windows often requires editing the registry, or installing third party add ons. Personnally, the cli is a lot easier for me to work with than the Windows Registry.

To address your "issues",
First: I guess you missed seeing the graphical frontend for your package manager (can't remember the name of the one for that distro)? You could have browsed a list of available packages (with descriptions and usually a link to the projects homepage, and installe them with a couple clicks of the mouse. Personnally, I find that much easier than driving to the Windows package manager (AKA Best Buy, Comp USA, etc).

Second: The reason your Wifi card works under Windows is because the maker of your laptop did all the dirty work (like driver installation and configuration) for you. Had you purchased the computer with Linux pre-installed the Wifi would have worked "out of the box" and you'ld be whining about having to make it work when you added the Windows install.

Third: same as above. Most laptops don't function correctly on the regular version of Windows, either - that's why the maker of your laptop provided the customized install and recovery disc (or partition). Chances are your WIFI and APM stuff wouldn't work "out of the box" with a regular retail version of Windows

Fourth: it's all in your head. I've never felt that Linux is "cramped" compared to windows at the same resolution. And I vastly preferr the Gimp over photoshop for what I do - mostly web and 3-d textures. If I was a photographer I'm sure I'd prefer Photoshop - but the Gimp's "make seamless" tool makes it so much more usefull than PS for me. That's just a matter of preference - YMMV.

As for compiling things - you don't need to with most distros. I used Linux successfully and happily for a couple years before I ever had reason to compile anything.

I think the only real problem you have with Linux is that you don't know how to use it. Once upon a time you didn't know how to use Windows, either, but you learned. Now, however, you're a big bad "power user" and your ego won't let you go back to being a noob and learning Linux. To bad - your loss.

Tommy

Re:Don't confuse the market segments.

[gordo3000]

I think you got your example wrong. Vendors like intel cooperate just the same with windows as they do with Linux, they release compiled drivers. Why someone who runs linux feels they should get more than that confounds me.

Really, its as long as open source zealots are not willing to coexist with closed source(even though most closed source is willing to go the other way) many Linux distributions will have a steeper learning curve.

Re:Don't confuse the market segments.

[Synbiosis]

Heh, this entire thread reminds me of a bash quote:

<dm> I discovered that you'd never get an answer to a problem from Linux Gurus by asking. You have to troll in order for someone to help you with a Linux problem.
<dm> For example, I didn't know how to find files by contents and the man pages were way too confusing. What did I do? I knew from experience that if I just asked, I'd be told to read the man pages even though it was too hard for me.
<dm> Instead, I did what works. Trolling. By stating that Linux sucked because it was so hard to find a file compared to Windows, I got every self-described Linux Guru around the world coming to my aid. They gave me examples after examples of different ways to do it. All this in order to prove to everyone that Linux was better.
* ion has quit IRC (Ping timeout)
<dm> brings a tear to my eye... :') so true..
<dm> So if you're starting out Linux, I advise you to use the same method as I did to get help. Start the sentence with "Linux is gay because it can't do XXX like Windows can". You will have PhDs running to tell you how to solve your problems.
<dm> this person must be a kindred spirit of mine

Re:Don't confuse the market segments.

[Randseed]

You aren't missing anything. These hardware vendors are in bed with Microsoft. Nothing keeps them from releasing the specs, letting someone else write a driver for it, and letting someone else do the technical support. And before someone says that then they'll be flooded with tech support calls, they already are from users who can't get their hardware to work, or are using some kludge piece of crap like ndiswrapper and having problems under it.

One machine I have has a DLink wireless card in it. I installed the acx_100 driver, which seems to be the driver for it. Unfortunately, the manufacturer has some bogus "feature" in the hardware that causes the network card to reset itself every so often. When that happens, the network card drops off the net, and doesn't come back up until a reboot.

The proper thing for the manufacturer to do is either release a Linux driver for their card, or document the incompatability so that I can hack support for it into the driver and release it. In this case, I've done the work for them, for crying out loud.

By the way, I did call the DLink cretins and ask about this. They knew nothing. Their answer was that they don't support Linux. When you release hardware, either release drivers for the predominant operating systems (which is now Linux and Windows), or at least have the decency to release the specs so someone can cover your lazy ass. They don't do this.

I think they honestly think that someone isn't eventually going to get pissed off enough to reverse engineer the damned card, which if I can ever isolate the problem I'm going to do and blow all over the Internet, along with a message proclaiming how much their company sucks.

BroadCom is another one. They release no information about their cards, and lock users into them by building them into the laptop. So you buy a new HP laptop, and its wireless card is a BroadCom. Oops. Sorry.

Re:Don't confuse the market segments.

[Vantage13]

IMO, if a driver for a wireless card that's as popular as the Intel ones still isn't supported in a default install when drivers were released months ago, Linux still has serious support issues.

So I suppose when XP doesn't come by default with drivers for hardware released two months ago that it too has serious support issues?

XP won't come with drivers for hardware that has come out after its release. The manufacturer provides the drivers. So if hardware manufacturers can provide drivers for XP is it unreasonable for them to produce Linux drivers on that same CD? On top of that they go to great effort to provide it in the format requested by MS themselves to be compatible with their OS.

Is it unreasonable for them to provide a linux driver in the format requested by Linux developers?

Is it reasonable to blame the Linux distro and not MS even though both haven't included the driver in the default install?

I don't see how these two situations are really different? XP needs third party driver for your card to work. Linux needs third party driver for your card to work. If anything, it's up to the hardware manufacturer to make sure the user experience is the best it can be. At the moment, they don't seem to be trying all that hard...

Re:Don't confuse the market segments.

[MightyMartian]

And yet I, a guy who does tech support for a small ISP, is going to get calls as to Norton's latest piece of crap firewall and virus package is blocking all web traffic, or not letting the guy send his email. His printer will suddenly stop working, and he'll have to pay some pimply 18 year old to get rid of all the spyware he's managed to get on his system. Even worse, if he goes out and gets a new video card, he may find out that the drivers aren't quite up to snuff and performance is the shits.

Windows is no panacea. If it was, there wouldn't be an entire tech support and repair industry based around it.

As to it's GUI, that really hasn't changed, despite cosmetics, in a decade (and is still inferior to the old OS/2 WPS that was badly ripped off for Chicago). There's nothing particularly marvelous about the GUI, and half the people I deal with suddenly find themselves in the deepest, uncharted jungle when you ask them to click on the Start button. It's probably one of the least innovative GUIs ever developed. What Windows has that other desktop-class OSs don't is about fifteen years of bullying and underhanded dealings with hardware manufacturers. It scammed, threatened and robbed its way to the top, so that now anyone with another operating system is hitting a brick wall whose sole purpose is to keep MS on top. If some of these hardware manufacturers would simply give the damn specs to the open source community, that advantage would disappear. By now KDE and Gnome are at the very least the equals of Windows, and I tend to think they may be slightly better.

So, in short, is Linux perfect? Nope, absolutely not. But then again, Windows isn't either, and the advantage that it has is due to its maker being a predatory monopolist, and nothing to do with any particular technical advantage.

Re:Don't confuse the market segments.

[IamTheRealMike]

That's correct, Ubuntu Universe is disabled by default, despite the fact that most of the packages you will want are there.

This is silly. It's been raised on the Ubuntu lists time and time again. Nonetheless, they refuse to enable it, because Canonical can't guarantee security updates for it. That's also silly, in my books, but then I find the whole idea of a Universe repository very silly indeed; it's simply not possible for a distro to ever be 'finished' like that.

Unfortunately, whilst new means of distributing software are being developed, the distribution developers generally strongly dislike them and sometimes go out of their way to cause difficulty for them. At some point there'll probably be a new wave of distros derived from the current ones that take the last "easy steps" to make Linux really usable for family and friends. Hell I'll do it myself one day if need be. Fedora is so close yet so far!

Still ... it's easy for Asa to criticise now. But think about all those years that Mozilla lumbered on with essentially zero popularity outside of the geek world. It wasn't until Firefox (which took years to develop) reached version 1 that things really took off. Linux is still in the Mozilla Seamonkey stage: cute features are being developed but that last stretch hasn't been reached yet. Arguably, it hasn't even been started.

Linux needs a Screen of Death!

[tktk]

Windows has the Blue Screen of Death.
Longhorn has the Red Screen of Death.
Mac OS 9 had the Sad Mac and the Bomb.
OS X has the big power button in the background. (And maybe one more.)

And Linux? I don't know what it's got.

Linux users need some iconic way to know that they've really fsck'ed up the their computer. Then they can be satified when they haven't seen it in a while.

We need something....like a dead penguin. Or maybe a slightly stunned penguin. I don't know, get a Japanese manga artist to draw it.

Perhaps the wrong approach

[Sv-Manowar]

The article seems to suggest that the general idea of "Putting things in the "right" place for Windows users will go a long way" is something that would be beneficial to linux switchers. The many users who have switched to OS X haven't needed this, and in fact have moved to systems where menu choices and design philosophy are significantly different to windows.

The reason for this not being a problem is that things are laid out in a way that's intuitive to those who just want to perform the action, rather than perform it in the way windows does. From my experience people who mostly use macs find it harder to use windows pc's on occasion than vice versa for precisely this reason. Windows has its usabilit nuances, and cloning them doesn't help people get a better experience from using the computer

Some good points, but...

[Coryoth]

First of all migration is raised as an issue: "When Regular People fire up the Linux desktop for the first time, the browser, office suite, email client, IM client, file manager, etc, each need to carry over as much as possible of the Windows application settings and all or very nearly all of the user data."

First of all that's a steep ask, but secondly I just don't think it's necessary. If that was required for people to switch no one would ever move to Apple. It's definitely a nice idea, and in the "nice to have" category, but I don't see that it's a deal-breaker.

The second point is API stability: "A user should be able to install Fedora Core 4 and go grab the latest Firefox release from Download.com and have it work without the need for finding and installing compat-libstdc++ or whatever."

This one is fixed - if developers would actually pay attention. Autopackage allows developers to package up their application into a self installing executable that can do dependency resolution. At that point not having compat-libstdc++ is the developer/packager's fault: they ought to have included an Autopackage for it in their repository so the installer can fetch it if it finds the right version of compat-libstdc++ isn't already installed. Better still, the people at Autopackage provide relaytool which allows developers to smoothly fallback to other library versions: for example, you can have your binary use the new GTK+ file chooser if it is available, but fallback to using the old one if it isn't. Which is really saying that the problem has been solved, it's up to the developers and people releasing the software to make use of the tools available.

The third point is preferences: "Gedit has about 30 user preferences spread across 5 tabs in a preferences window -- Notepad has about three."

Now that's not a great example becaue Gedit does a hell of a lot more than notepad, but I think the point is still very valid. To be fair I think GNOME has been putting in a lot of work on this front, and trying to clean a lot of these things up. That work is ongoing, and we can expect to see continuing improvment. That is, the way forward has been laid out, it's just a matter of continuing down the path.

The final point is "comfort":"The final major issue is comfort. Linux must feel comfortable to Windows users. Most people using computers today have been at it for a while now and they've been at it on Windows. Don't mess with their basic understanding of how things work."

I have to say, I think this one is a little dubious. If there is a better way of doing things why not do it? I think constraining yourself to the way Windows does things is a little pointless. There are plenty of things Windows does well, and it's fine to follow those examples, but there are plenty of things Windows does badly, and slavishly copying broken behaviour really doesn't make much sense.

I think the real point here is: be patient. I think the points are valid, but they are also largely well known, and being dealt with. Linux on the desktop is not going to "take off" anytime soon, but the rate of improvment in desktop Linux is tremendous, and it is making slow but steady inraods. Software installation (which has been the recent bugbear that people complain about) is looking quite good with Autopackage and Smart, but both of those are very new and it's going to take some time before a lot of stuff shifts over - that's life. GNOME is working hard on the preferences trim down and clean up, and, I think, is workign towards a fairly clean easy to use Desktop. KDE is headed in a different, but equally valid and interesting direction - I think the divergence is going to end up providing some real significant choice. Finally I think once all these bits properly fall into place and desktop Linux manages to make a dent in the enterprise (which seems to be where the major distros

Re:Some good points, but...

[asa]

f that was required for people to switch no one would ever move to Apple.

This is the third or forth time I've seem mention of this on this thread. I think you all overestimate the number of people switching to Apple. How many users have they actually taken from Windows? A million? Two million? It certainly isn't much more than that. I know it's apples to oranges, but Firefox gets about that many IE switchers _every_week_. How did we go from one or two million users total just a couple of years ago to many tens of millions of users today? In part by making it work the way IE users expected it to work (without dropping innovative and powerful features like tabbed browsing and pop-up blocking.)

If there is a better way of doing things why not do it? I think constraining yourself to the way Windows does things is a little pointless.

Why not do it? Because users don't like change. Because sometimes habit and comfort are more important than making it marginally better. Unless you can make a dramatic value addition for the user, change is probably a bad idea.

I think desktop Linux is looking good, and it's just a matter of time before it manages to carve out its niche.

I'm a big fan of Linux (and Mac and Windows, too) and I want it to be successful. I didn't write a blog post saying "Linux can't and won't succeed," I tried to point out the areas that I think are conditions for its success. I think there's a big opportunity pre-longhorn to show that we've got the right stuff. That window is closing and things will be much more difficult after that. I think it's dangerous for us to think "it's just a matter of time."

- A

if you don't help newbies...

[yagu]

From the post:

needs to do to get there for the "regular user" AKA mom, dad and grandma Bootsie

(First off, I'm a little nervous about how the OP knew my grandma's name.)

If you don't help newbies with linux, especially ones not very technical, then linux may not be ready for mom, dad, and grandma. Applying this standard implies also then Windows is not ready for mom, dad, and grandma. I've spent countless hours (that I can't charge, and I'll NEVER get back) fixing, re-installing, helping, instructing, etc. in a support role for my parents from Windows 95 through Windows XP.

And, guess what? They're still struggling. Part of this stems from the fact they missed the technical revolution (and lest you diss my parents, one is a Doctor, the other is a Concert Violinist, played in the Pittsburgh Symphony). But most of it stems from the intractable problem of rendering technology intuitive and transparent to the lay-person.

Interestingly this problem plagues both Windows and linux. Interestingly, for Windows what I've found in coaxing my parents along the learning curve is Microsoft has done much if not most to make Windows obfuscated to my parents. Each new generation has left them re-learning pieces of the environment they had just about almost mastered... (they were this close!)

But, I do think linux is up to the desktop task for many who use the internet for mostly surfing, e-mail, quick word docs, and simple spreadsheets. And I think linux actually fares better simply for the rock solid reliability. I haven't set up my parents with linux because I live 2000 miles away from them, so I'm a little paranoid that should something really bizarre happen, I wouldn't know who to have help them, while with Windows, though it demands more support, if I'm not available, there's always some quasi-pseudo expert ready to jump in and "fix" things.

However I have set up others with linux, and I've been amazed... the support calls simply stop! This is for people who satisfy the above criteria: internet surfers; e-mail junkies; and simple "office" tasks. The linux just works. There's probably a larger demographic out there that could use linux than most people think.

A fine post and everything but...

[flithm]

how the hell did you get first post with so much freakin' writing?

Did you type this up ages ago, just waiting for the day you could get first post with this?

As for my comments on your fine post:

I recently installed Linux on my non-computer literate girlfriend's computer. She was always afraid of Linux because whenever she tried to use my computer it was nothing like Windows.

I put on a nice easy to use distro, set her up with KDE, and let her go to town. She's now using GIMP, uses it for all her photographic needs (scanning, digital camera). She even sighs when she has to reboot to Windows.

She was amazed at the little things, like how cut 'n paste works. I could tell she was thinking "why wasn't it always this easy?"

Or how she can resize an entire "folder" of images with a couple of clicks and no fuss.

I really don't think it's so much a matter of Linux offering nothing of value that Windows doesn't, because that's simply not true. Linux has tons to offer the average person that Windows doesn't.

The major problem, as I see it, is that it requires changing the way you think about using a computer.

When I first started using Linux I got very frustrated for a while, simply because my mind is notoriously bad for resisting change. It didn't like having to re learn such simple stuff. In fact in the beginning I kind of felt like I was a prisoner to my computer. I no longer knew how it worked at all. No idea! How do things run at startup? How do I add a printer? It was all this huge mystery.

And then, even beyond that, everything is just Done Differently. You really have to change your mindset to become a fully functioning *nix/*BSD user.

For a lot of people that's a really hard thing to do. But the funny thing is it really doesn't take that long. No longer than a week later my girlfriend was installing her own applications, updating her system, etc.

Anyway I don't want to give the impression I don't agree with what you said, because that really was a good and well thought out post. For the most part I agree with what you said, I just wanted to add that.

Re:A fine post and everything but...

[Sloppy]

how the hell did you get first post with so much freakin' writing?

That post is one of TMP's recent journal entries.

BTW, TMP is one of my favorite ranters on Slashdot, but this one .. I don't know. It just isn't angry enough. I wish some Linux-using republican would tailgate The Mad Poster for ten minutes, then pass him on the right-side shoulder while going 90 mph, revealing a bumper sticker on the back of their SUV, showing a penguin flipping him off. Then maybe we'd get a better rant.

C'mon, TMP, lay off the prozac.

I don't think Windows is desktop ready

[Nailer]

• You install it, there's no apps (or crap ones - compare IE to Firefox or Outlook Express to Evolution), or you pay lots of money to get them.
  
• You run as root by default, not for ease of use (how difficult is 'type your passsword to continue' that Fedora and OS X do?) but because Microsoft and Windows developers couldn't be bothered fixing things. And you get spyware and viruses as a result, and you fix most, but a few remain, and the whole thing moves like molasses. The SP2 'firewall' still lets in about 7 network ports by default, including those used for some of the major worms.
  

That's why I recommend Linux. I don't see either item changing soon either. I've played with Longhorns betas, and nothing's different. It's your computer, you may as well use it. That's why I recommend Fedora.

Re:less stupid users

[tomstdenis]

That's like saying the house you build that won't stand a 10km/h wind shows how "construction work is easy".

If windows were easy to use it would be easy to use securely.

The fact that grandpa can't install, update, manage, us the anti-{spy, mal, virus, windows} software you have to install shows that it's not easy to use.

Next you'll say open-heart surgery is easy as well....

Tom

Re:We?

[snorklewacker]

I've been using Linux fulltime on the desktop since 1999-2000. What pisses me off is applications switching between ok/cancel positions themselves. When I don't need to worry about where the OK button is going to pop up in Firefox/Mozilla then I'll start to worry about the rest of the OE.

You can thank gnome for that. They decided that since That Other Desktop Environment was ordering buttons according to the Windows interface convention (of "ok/cancel", "yes/no"), they'd just switch to the mac convention of "cancel/ok", "no/yes". They trotted out some high minded theory about how the lower-right corner of the dialog was "special", and how all these HCI studies (all put on by Apple of course) proved this was an enlightened change ... but as usual for gnome these days, it was just another gratuitous jarring change. End users were told to suck it up and bask in the glow of gnome's superior wisdom.

So that's why they're backward.

You just proved my point.

[khasim]

You claim to be a power user, but you had problems with your wireless card and power management.

I said that power users would be the last segment (#4) to move to Linux because they wouldn't be happy until their hardware was supported.

I run Ubuntu and it runs great on fully supported hardware. But then, I also run my LCD screen as 1280x1024 so I don't have the space problems you do.

Re:You just proved my point.

[the_womble]

Not only will "power users" be the last to use Windows because they want all their hardware supported, but they are also usually specially "windows power users" they have invested a lot in learning how Windows and whatever apps they use do things, but they do not actually understand how they work so their "knowledge" is not transferable.

This is actually the group who the article call "regular users", real regular users are quite happy with Linux desktops - copy their files over, export their bookmarks and import into Firefox and that's it. This has worked fine for my father, my wife and some guys who worked for me (one is now planning to install Linux at home).

I also do not understand what he is talking about when it comes to installing applications. There are only three pieces of software I have installed which required anything more complicated than downloading the RPM, clicking in it to start the installer, and then typing the root password and clicking OK a few times. These were: Erlang, Firefox and Thunderbird.

In fact, bar Erlang (which needed to be compiled), Firefox has been by for the most problematic thing to install.

Re:You just proved my point.

[the_womble]

I'm not sure how you're able to make the claim that Windows power users only know how to do things, but don't know how it actually works.

I make the claim because I have known people for whom it is true. "Power users" known recipes for getting things done, and secondly their knowledge tends to be very narrow.

I think that you do not understand how people who know absolutely nothing about computers approach them. I would have agreed with your statement at one time, but I have slowly realised how many people get a lot of stuff done by learning sequences of actions, rather than actually understanding what is going on. Yes they do inevitably learn a little (especially if they start writing macros), but it is much less than you might think. If you have absolutely no idea of how computers work, you have no framework to learn from. A computer becomes a black box device that produces certain outputs for certain inputs and that's it.

Most people do not actually do much configuration beyond installing software (which these days is easy), and setting backgrounds and screen savers (and even there many users call the former the latter).

As of the narrowness of power users knowledge, let me give you a few examples. Many years ago I came across someone keeping a database in Wordperfect. They knew WordPerfect so they wrote a set of macros to do what they needed. That is a power user in action. More recently I have seen Excel used to circulate information - so that in order to see a single page that you wanted, you had to download an Excel file that ran to several megabyte with macros etc., the file had to be manually copied to the file server at each branch office. Putting the information on a web server would have been obviously better. This was the product of a "power user" who knew how to write VB scripts in Excel but little else.

As for software installation, I have not used OS X , but I would say that the better Linux distros (such as Mandrake) are at least as easy as Windows - easier if you stick to software from your distro. The hardest are of course very difficult to install (both OS and additional software), but they are designed for a different user base.

Re:You just proved my point.

[indifferent children]

I have to say, by far the easiest operating system I've ever installed anything on is OS X. Just open the disc image and basically drag the file onto your desktop or wherever and boom! It (usually) works! After that, I feel that Windows is the next easiest (that might have to do with the fact that Windows programs are so ubiquitous and therefore easy to find... and you usually have all the DLLs and such from previous installs that actually shipped with them. Then again, this leads to "DLL-Hell"...) followed by Linux as being the most difficult to install things on, on average.

I don't think that you have used any of the Debian-based distros. The Debian apt-get is fantastic (the RPM-based apt-gets are still inferior). You need to know the name of the program that you want to install. After that, it is: apt-get install mozilla-firefox . And apt-get resolves all DLL-Hell-ish dependencies for you (recursively).

If you prefer GUIs, 'synaptic' shows you lists of all known applications, you click the apps that you want, and synaptic runs apt-get for you.

With OSX and Win32, you must go out on the Internet and find the packages that you want to install, download them, download any other packages to satisfy dependencies, and then use the 'easy' install procedures (in the correct dependency-driven order). Apt-get wins hands-down.

Asa is right

[jalefkowit]

Unsurprisingly there's already a lot of "bah, this guy wants Linux dumbed down for n00bs" comments on this thread. Which totally misses the point:

Linux-on-the-desktop isn't just too complicated for n00bs -- it's too complicated for reasonably sharp users, too. And that's the problem.

I offer myself as an example. I am not the God of All Things Computing. But I've been tinkering with PCs since MS-DOS 3 days, I've used Windows, Macs, Linux and even CP/M for pete's sake. Today my primary desktop at home runs Ubuntu Linux. I'm comfortable compiling software from source tarballs and rooting through Google for HOWTOs and FAQs.

In short, I know my way around a computer -- and yet Ubuntu still manages to throw me for a loop more frequently than I'd like.

Example. The other day I installed the new Deer Park preview of Firefox. For some reason, its installer (bonus points to it for even having a graphical installer, btw) didn't add a shortcut for launching it to my GNOME panel. So I wanted to add one myself.

Easy? Right? Bzzt.

On Windows, here's the steps for adding a new item to the Start menu:

1. Click Start menu button
2. Navigate to folder where you'd like to add shortcut
3. Right-click folder name
4. Select "New Shortcut"
5. Wizard launches that walks you through finding the program you want the shortcut to point to, and giving the shortcut a name.

I figured there must be a way to manipulate the GNOME panel in a similar fashion. Nope. There is no direct way in Ubuntu Hoary to add a panel item to the menus through the GUI. Instead you have to open a shell, find /usr/share/applications, and create a .desktop file in there for your application.

But! You don't have permission to do that by default, so you have to use sudo to create the file. ("You do know how to use sudo, right Mom?")

And then -- once you figure out that you need to create a .desktop file, and where this file needs to go, and what format this file needs to be in, and you actually go and create it -- nothing happens! That's right, you don't see the item in your panel until the next time you log in, unless you manually restart the X server with CTRL-ALT-BACKSPACE.

(Yes, you have to restart the window manager, or else it will appear that all your work was for naught. "Just restart the X server, Mom. Mom? Hello? Noob.")

The icing on the cake is that to find this answer, you have to go through three levels of redirection:

• Ubuntu tells you to refer to GNOME, since it's their desktop, Ubuntu's just distributing it
• GNOME tells you to refer to FreeDesktop.org, since it's their standard for panel items, GNOME is just packaging it
• FreeDesktop.org hides the instructions on how to write a .desktop file deep in a standards document.

("You do read standards documents, right Mom?")

I went through all that and finally got my shortcut added to the panel. But how many average users are gonna put up with that? (And Ubuntu does better at this stuff than most others.)

With all the spit and polish issues that Linux has, Asa is not the only Mozillian to find fault with it; former Moz UI gadfly Matthew Thomas (aka mpt), who's now with Ubuntu sponsor Canonical, recently posted a list of 69 usability flaws in Ubuntu Hoary, and old skooler Jamie Zawinski gave up Linux for OS X for good.

My case was not a case of "user who could not snap out of Windows-ism". I'm more than willing to embrace a better approach when I see it. But this is not a better approach fo

You forgot the the biggest segment

[EmbeddedJanitor]

Embedded. Cell phones, TVs, dektop boxes.

Linux excells where Joe Sixpack does not have to fiddle with set up. That includes situations where the computer is not visible to the users (embedded and servers) as well as those where someone else completely manages the box (eg. corporate desktops).

For the general home user I agree that Linux is a pig. I can't get my PC to play MP3s. The Winmodem needed a bunnch of hacking etc.

Re:Tired of the moaning

[natrius]

Everybody seems to think "well just because it doesn't work like Windows then it is flawed."

The article wasn't supposed to be a fair analysis of Linux. New users don't give thing fair analyses. If they don't like the button order and don't want to adapt, they won't use it. That was his point.

Re:instead of linux it must be...

[linguae]

KDE, GNOME, XFCE this has nothing to do with linux, it affects BSDs too.

And, frankly, that's the biggest problem Linux faces toward desktop acceptance. No, forget Linux. That's the biggest problem Unix faces toward desktop acceptance, whether it is Linux, BSD, Solaris, AIX, HP-UX, etc. Unix has been dependent on the X Window System for its non-console output needs. For those of you who don't know the story, let me tell the story.

The people over X didn't agree on a single toolkit that all X applications use, and left that work to other developers. Motif was the official X toolkit for one point, but Motif was closed source and therefore not embraced by BSD and Linux developers, not to mention Motif is ugly. LessTif (a Motif clone) was developed to allow Motif developers to compile/run their apps in Linux, and some projects (notable the GIMP), created their own toolkits. Years later, a Linux user comes along and says, "Motif is ugly and hard to code in. CDE isn't very user friendly, and they're both proprietary. Imagine if we had a toolkit that was easy to develop with, and a desktop environment that's easy to use, just like Windows/Macintosh. I'll call it KDE!" So this developer starts work on KDE, which uses the QT toolkit.

Unfortunately, QT was released under a license that wasn't compatible with the GNU GPL, which is one of the most common licenses used on OSS development. So, instead of the GNU folks writing a GPL-licensed QT clone so that way they can still ethically use the growing amount of KDE applications out there, they decided to write their own toolkit and their own desktop environment. They adopted Gimp's toolkit (GTK) and started work on GNOME. Before long, users now had two competing desktop environments (KDE and GNOME), multitudes of toolkits (Motif [which became open-sourced 5 years ago], QT [which was GPL'd for OSS projects], GTK, wxWindows, GNUstep), and vast amounts of software tied to one toolkit.

The problem with getting Linux on the desktop is that even though a "Joe Average-ready" distribution (like Mandrake, Linspire, Ubuntu, or Fedora) picks a desktop environment, picks the applications (which depends on the desktop environment), and develops its own installer, we all call them "Linux," as if you can go down to the store and buy yourself a box of Linux 2.6.11. The problem with that is each distribution should be really regarded as its own OS, since each distribution may use a different desktop environment. For example, we don't call Mac OS X "FreeBSD," even though much of the underlying OS is based on FreeBSD. However, we don't refer to Mandrake or Ubuntu as "Mandrake OS" or "Ubuntu OS." The biggest problem with this all is when the user tries to install software. Let's say that the user decides to install Ubuntu on his computer. IIRC, Ubuntu uses GNOME as its main desktop environment. The user doesn't know about GTK, GNOME, Bonobo, ATK, Gail, and all of that other technical mess. However, let's say that the user has heard of a wonderful application called KOffice which meets his needs perfectly. When the user installs KOffice, he finds out that he needs to install another toolkit (all of that KDE/QT stuff) in order to run that application. When he opens KWord, he notices that everything from the buttons, the menus, and even the ordering of the "OK" and "Cancel" buttons are different than from a "native" application such as Evolution or GEdit. Being ignorant about toolkits, he then installs some other applications such as xpdf, some GTK 1.2 applications, OpenOffice (which uses its own toolkit), some Java Swing applications, and more. By the time he's finished, he would have to deal with almost a separate toolkit for each application that he's using.

The biggest problem with the Unix desktop is the X Window System. There is no consistency with the look-and-feel of applications. There is also no consistency with toolkits. Mac OS X did it right with Aqua, but Aqua isn't the X Window S

Re:Tired of the moaning

[asa]

Everybody seems to think "well just because it doesn't work like Windows then it is flawed." We should not (and will not) bow down to these kinds of gripes. The coummunity is in the business of producing better software--not equal software.

I don't think "just becuase it doesn't work like Windows then it is flawed" and I didn't say anything like that in my blog post. I said that if you want to get Windows users to migrate to Linux, you need to make the transition as easy as possible and that often means making some features and behaviors work like Windows.

Firefox didn't adopt IE's "be overrun by pop-ups" feature but we did adopt Ald+D to focus the addressbar. We decided it was worth more to the user to give them a pop-up free browser than to try to train them to use Ctrl+L to focus the addressbar.

It's not one or the other. Pick your battles. For linux to be successful in converting Windows users, it is going to have to make smart decisions about these kinds of issues. I can see approximately zero value in reversing the OK and Cancel buttons and I can see it being a very uncomfortable re-learning curve with a lot of pain when the user gets it wrong out of habit. Where is the value there. Why throw up that difficulty.

- A

Disappointed

[RAMMS EIN]

The scoop made me think that this would be a nice comparison of what made Firefox get adopted so quickly, and Linux so slowly. Instead, it gives the same old crappy arguments of why Linux is "not ready":

``The first issue, migration, is pretty serious.''

No, it's not. You don't need to run it next to Windows. You don't need to provide the same applications. You don't even need to provide an equivalent for every app. Nor all the games. OS X doesn't have all this. Is OS X not ready for the desktop?

``The second problem that blocks massive Linux Desktop growth is stability.'' (The use of "stability" is confusing. What he means is: you can go to a website, download an application, and expect it to run, i.e. binary compatibility)

This is the Windows Way. Linux has a better alternative: packaging. Applications packaged, tailored, and tested for your distribution. Try Debian. Go through a number of installs, uninstalls, upgrades, and dist-upgrades. Then tell me if you like the Windows Way better.

If you do like the Windows Way better, there is hope for you. It _is_ actually possible to distribute binaries that work. Opera has been doing it. StarOffice did this last time I checked (a very long time ago). I'm sure there are others.

``The third issue is a lack simplicity.''

The complaint here is that Linux gives you too much choice. Choice is not an antonym of simplicity. Try Ubuntu. Installation requires that you select a drive to install on, create a user account, select your keyboard and timezone, and wait for stuff to install. No hard choices there. Once installed, it has a nice GUI environment with one app for every job. Just because the choices exist, doesn't mean _you_ have to face them. You can have other people make them for you.

All the 237584704908c34 window managers are for people who like to experiment and try new things. If you don't want to bother with them, then don't.

``The final major issue is comfort. Linux must feel comfortable to Windows users.''

AKA, everything needs to be called the same and be in the exact same place as on Windows. Again, see the earlier argument about OS X. As for the new concepts of mounting and unmounting, have you heard of automount? I believe KNOPPIX uses it, complete with icons appearing on the desktop when you insert a drive.

So, with all these issues declared junk, what do I think is holding back Linux? Here's my list:

1. Linux isn't shoved down people's throats. This is why people have to "switch" in the first place. When people start using computers, they run Windows. At work, computers run Windows. When you buy a computer, it has Windows installed. Sure, there are exceptions, but for all practical purposes computer = Windows.

2. People don't care. Many in the Linux community want people to switch to this "better" system. To most people, Windows works fine. Why fix what isn't broken? This is also why Firefox users are still outnumbered by MSIE users.

3. The issues you raised are widely _seen_ as problems by people who haven't actually used Linux. Linux has a bad reputation for being user-unfriendly, which is entirely undeserved (and has been for years). One could even argue that the security problems with Windows make Linux easier for non-experts.

4. People are not sufficiently aware. They are not aware of how bad Windows is. They are not aware of how good Linux is. They are often not even aware that there is an alternative (they may have heard of Linux, but not understand what it is). If we want more users to switch, we need to educate people.

As for me, I don't really care what other people use. I've used DOS, Windows, various Linux distros, NetBSD, OpenBSD, Mac OS, and OS X. I like how I can write a program on one of the unix systems, then compile and run it on another. I don't like that it won't work on Windows, and that Windows is missing so many basic things, but Cygwin goes a

Re:Exactly!

[It'sYerMam]

While a universal type of package might be useful, it's probably unlikely. What's more likely, I expect, is a source "package," like the emerge system except universal. You download, say, "firefox.ins" which is a tar.gz/bz2 file including at the root something like an INSTALL file, except designed to be read by a package manager. It could just be a shell script, or something more like an xml file including instructions to hand to make, ./configure and so on. When to ask for root passwords, perhaps define some kind of GUI for options to pass to the configure script.

Basically an installshield equivalent, that then becomes capable of installing on any distro. Sure, you still have the speed of installation being an issue what with compilation, but hopefully a system like this will be available in the not-too-distant future.