Linux in 2004?

An anonymous reader writes "John Terpstra and Eric S. Raymond have started the ball rolling on LinuxWorld's poll of the community for what they think will happen in the world of Linux in 2004. Terpstra says 'I predict that during 2004 at least one significant USA government body will adopt Linux on the desktop.'" Depending on how you define "significant", this has already occurred.

Comments

Apt

[rf0]

I found the following intresting

"I think 2004 is going to be a big year for Fedora and Suse, and a challenge for Debian (because Fedora now offers apt for RPM)."

Well apart from the fact that apt for rpm has been around for a while and also debian packages usually come configured a lot better than fedora are aren't as buggy.

Of course with the recent Debian security breach things might not be that easy

Rus

Doesn't DOD already use it?

[Anonymous Coward]

I thought that DOD went out and purchased 10000 licenses of Star Office from Sun... Weren't those for Linux? Or were they for winders?

The image we want to project?

[toddestan]

Maybe having Linux being "good enough for government work" isn't exactly the image we want Linux to have. Just like I think having Linux on cheap, disposable, sub-par computers from places like Wal-mart may not be the best thing either.

The real goal is to have people see Linux as a viable alternative, not a cheap Windows imitation or some eccentric thing the government uses.

Re:The image we want to project?

[Prof.Phreak]

VERY good point. I've heard so many stories about "Oh, I've tried Linux and it's slow", and then questioning further, I find out that they've installed it on their old Pentium1 box - well D0H, of course it will be slow compared to Windows running on a P4!

The image needs to be: If you have a GOOD computer in your home, use it for Linux! Give the trash computer to Windows. Most people will quickly realize that Linux outperforms Windows and is snappier to use if the hardware is equivalent.

(I actually have 2 computers with relatively common hardware, one running XP, the other running Linux, and I must say that Linux feels much faster).

Re:The image we want to project?

[GreyWolf3000]

The software quality already speaks for itself. Nobody uses Linux.

Actually, across the world millions of people use Linux.

Why? Because it supports a tiny fraction of the hardware normal people actually want to use,

Define "tiny." It actually supports more hardware than Windows does. It also supports most hardware that most people have. Hardly "tiny."

because other than Mozilla, Evolution and Open Office there is simply no usable software that runs on it,

I use many, many more pieces of software than those three.

because the distros come with 100s of completely awful, useless piece of crap betas with no sense of a homogeneous user interface,

Linux distributions offer more software by default than Windows does. There is crap software for Windows as well. If Windows shipped with, say, a really lame filesharing client that constantly crashed, would that have anything to do with the quality of the software that said client is running on?

because at least 3 different modern distros actually don't support dual monitors on 2 separate video cards in spite of what the config software says,

The distros all run the same software, hence if something works on one distribution, it can run on the rest of them. Some distributions automatically take care of it, however, and others do not. Whether or not the end user should be responsible for configuring his system manually is another matter entirely. In any case, your claim is just plain wrong.

because the config is so moronic that it will let you do things that render your computer completely impossible to use without a warning

Why, specifically, should Linux be on trial for the mistakes of the user? Where did Linux claim itself to be responsible for the actions of it's users? What part of "use at your own risk" do you not get? What makes you think you cannot destroy a Windows setup with equal ease? Hint: you can. The "safeguards" are a paper-thin veneer.

(case in point: I set up dual monitors, but the GUI config utility didn't have any option to set the layout, i.e. 2nd monitor left of 1st monitor or the opposite, and so when I rebooted the main screen was shifted over to the right outside of the display area of the rightmost monitor, which meant I had no display; spending 3 hours online with a Linux support guy didn't manage to get dual monitors to work by hand-editing the X config file, even though the OS recognized both cards and both monitors by name; the OS thought that monitor 1 was hooked up to video card 2 and vice versa; etc).

I am not going to apologize for you not knowing how to set up your system properly. Distributions that claim to be as easy to use as Windows have a responsibility to make such matters work "seamlessly," but just because they make mistakes today doesn't mean that they will never be able to fix them.

In other words, because Linux is more like a 1952 Chevy truck than a 2003 BMW (or even a 1996 Toyota Corolla),

Doesn't seem that way to me.

in spite of what those BILL GATE$ I$ EVUL LINUX RULE$ morons will have you believe.

What about those WINDOW$ RULEZ LINUX SUCK$ morons?

In 1998, I heard "give Linux a couple of years to catch up and it will solve all your problems". It's almost 2004 and Linux is still crap, a glorified beta that inept geeks will try to have you believe is the best thing since slice bread. Sheesh.

Something tells me you haven't tried it recently. If you think that Linux has not made progress since 1998, then you are either a) completely unknowledgable, or b) a complete and utter moron.

Please tell me you're just misinformed.

Linux is not a product. It's a process. And it works.

Moving that way

[ImTwoSlick]

I work at a USGS mapping center, and about 75% of us programmers use Linux almost exclusively. The IT department also has quite a few linux servers running too.

The real influx of Linux is due to the hiring of university students. Push Linux in the schools, and it'll end up in businesses and the gov.

US Gov't Agency Linux Desktops?

[DynaSoar]

Terpstra says 'I predict that during 2004 at least one significant USA government body will adopt Linux on the desktop.'

Not with Homeland Security showing how absolutely retarded they insist on being and going with WinXX. This is clearly not a security based decision, and any "significant" attempt to go counter to it will bring the HLS pseudo-spooks down by the thousands to protect their investments ^H^H^H^H^H^H^H^H^H^H^H The Nation.

michael sez: 'Depending on how you define "significant", this has already occurred.'

Pray tell, what agency might that be? In my years inside the beltway (up through less than 2 months ago) I didn't see any with any appreciable (let's define that as, say 5%) Linux desktops on desks. All I've seen, besides individuals setting up their own for number crunching, is piles and miles of MS systems "supported" by clue-deficient federal employess constantly in fear of replacement by contractors for extremely good reasons. Even NIH was mostly MS on the desks, and what wasn't was Macs. The necessarily more powerful research machines we used were often *nix, but these were not desktop machines.

Offering a secured version of Linux for D/L is not the same as an agency's internal deployment of same.

Nah, Education is the Future

[Qweezle]

Lower-education will be the short-term future of linux in my opinion. It's already hit the corporate server level, the governmental level, the ranks of high education, now what?

I've suggested to our [poor] school district that we should switch to Linux, using the old hardware we have, and they liked they idea but said it would be "too hard to implement". Oh, come now. I think that any kid could easily circumnavigate a Linux interface, especially if it is an easy one, like Mandrake 9.x or Lycoris! I sure would want my kids to learn Linux, and this is a cheap(free, actually) solution for those school districts that just can't seem to raise any money. In addition, get a good IT guy at the helm, fire all the low-waged IT guys who don't know what they're doing, and get that network running smoother than ever with Linux!

It's a SHOCK to me that school districts haven't at least started putting in an "Operating Systems 101" class in high school for everyone to learn about alternative OS's. Linux, Macintosh OS, Solaris, FreeBSD, UNIX, just imagine how much that would open up the minds of those kids!

Re:Nah, Education is the Future

[cscx]

Here's a reality check:

*All* new Linux distributions (e.g., the Mandrake 9.x or Lycoris) will run like shit on shit hardware. There seems to be some sort of misconception in the Linux community (psychologists call this "denial") in which its consistuents seem to think that (graphical) Linux will always run good on crap hardware -- in fact, it will run 908345 times faster than Windows!

Wrong.

At my university, I'm forced to use a Linux lab for one of my classes... unfortunately this lab is not blessed with brand new computers like the other 99.98% of computer labs on campus. No, these machines are Pentium II-400s with 128 MB of RAM. They are running RedHat 9 (kernel 2.4.20). Not too shabby, eh?

That is, till you log in to GNOME or KDE. It's slowdown city. We are talking I-can-sometimes-watch-windows-paint-across-the-scr een slow. Try and start up Mozilla? It takes about 20 seconds. No joke. Cripes, even gnome-terminal or Konsole take at least 5 seconds to start up. And in the swap meantime swap all the swap swap machine is doing is swapping to disk. Swap swap swap (get the point? ;) This is unacceptable. Linux needs a fast, NON-resource hogging, non-skinnable desktop that doesn't lack in the functionality area. Think iceWM, except not as half-assed when it comes to features.

Many Linux zealots like to talk about Windows bloat -- meanwhile there is precipitously more bloat in Linux software! (Kind of like that MS Office == bloat FUD that you see around here a lot. Just look at OpenOffice.org -- HAH!)

So what options does this leave you with? You obviously can't use KDE or GNOME on old hardware. Use twm? Yeah, right.

Win NT 4.0 still comes in handy for situations like these. I had to do an install recently where a machine had to be available 24/7 yet run on crap hardware. NT4 has an extremely fast, functional desktop, and it will easily install in It's a SHOCK to me that school districts haven't at least started putting in an "Operating Systems 101" class in high school for everyone to learn about alternative OS's. Linux, Macintosh OS, Solaris, FreeBSD, UNIX, just imagine how much that would open up the minds of those kids!

Well, here is some more shocking information: 98% of high school students don't give a rat's ass. Those that do already know. And there is a good number of school districts that have standardized on MacOS. Why do high schoolers have to learn about Solaris, etc? We're talking about people that have trouble grasping the concept of MS Word, so I think appreciating the benefits of Korn Shell can be left for some other time.

Unfortunately you seem to be a bit short sighted (evidenced by your association of "IT guys who don't know what they're doing" with "not using Linux") so this suggestion may be of limited use to you...

Re:Nah, Education is the Future

[Daengbo]

This is the reason that many K12 institutions in the northwest are turning to K12LTSP. (Re)use the old hardware, spend some of the new budget on a few servers, hook up the network, and you're finished.
As for the children's response in my school -- there is virtually none. Our children pick up the mouse, ask us where to go, and no other training is often necessary. They know precious little about computers, anyway. Eventually, we hear them bragging to their friends about being able to choose which desktop they want to use.

Re:What will drive Linux adoption

[Enucite]

Two good examples are RealMedia and Flash. I realise that there are solutions to both of these, but the quality is nothing compared to what is available for Mac or Windows.

Really? My experiences don't agree with your observations.

I haven't had any problems with Flash, it's just as annoying as the Windows version when displaying ads, lets me play the little flash games, navigate all the flash sites, and see all the flashtastic content on the web.

As for RealMedia, their new Helix Player has been working great for me. So good that I actually install it on all my Linux desktops, unlike RealOne for Windows which I never touch.

Always "a couple years away"

[Micah]

It seems like Linux on the desktop for the masses is always a couple years away.

In 1998, I was swearing up and down that by the beginning of 2000, some major PC manufacturer would be selling Linux-preloaded systems branded for comsumers in places like CompUSA. That obviously didn't happen.

For most of the last four years, I've been predicting that by the beginning of 2005, most people would be using open source operating systems (keeping in mind that that could be Windows, if Microsoft caught a clue in time). Doesn't look like that's going to happen.

Now it's looking to me like the first half of 2006 is when Linux use on the consumer desktop will move from the "early adopter" to the "early majority" phase. I say this because:

* It's virtually guaranteed that we'll have several more major deployments in 2004 and 2005. These might be specialized applications instead of general desktop, but that will help create demand for more general applications.

* If you read the "Roadmap to desktop Linux" posted earlier today, it's clear that several very cool and useful features will be coming to the Free desktop in the next couple years.

* OpenOffice.org 2.0 should be released in the first half of 2005, and it is planned to make development of add-ons much easier. This will hopefully help get more office-oriented vertical applications ported to OOo.

When all this happens over the next couple years, I believe desktop Linux will turn from a stream to an avalanche.

But still, we need consumer pre-loads with all hardware configured to work out of the box, and marketed well. Few people are going to buy a Windows-infested PC, then choose to replace Windows with Linux. This is probably the most iffy condition, but I think it will happen. Most PC manufacturers would do anything to break away from MSFT.

Re:Always "a couple years away"

[kris]

It seems like Linux on the desktop for the masses is always a couple years away.

Depends on who you are looking at. For me and about 2/3 of my colleagues, Linux on the desktop has already happened several years ago. I, being a consultant, am running Suse Linux on my laptop, on my business desktop, and on all machines at home.

I own a copy of VMware, but reviewing my usage of it, I only use it with Win98 to program my PBX at home, and most of the time with Linux to simulate certain customer configurations and experiment with RAID and cluster setups. All office work, including text processing, presentation and calculation is being done on Linux natively, as is web browsing, other internet work, and of couse all security work.

Kristian

Re:What will drive Linux adoption

[darnok]

Actually, I'd like to think 2004 might be the year Linux gets a feature-for-feature equivalent to MS Exchange, that supports MS Outlook, Evolution and a few other key clients.

When such a solution appears, that will mark a major milestone for Linux in potentially replacing Windows in many organisations.

Re:Two things that need to happen in 2004

[rsilvergun]

Um, why does the installer need to be easy? Most people get there computers from an OEM (Dell, Gateway, etc), who'll just include a restore disk anyways. There's this notion Linux will gain lots of market share with a grass roots campaign of installs on old Windows hardware. I don't see that happening (hasn't worked with Mozilla, and that's a much easier switch). On the other hand, MS is gonna burn a lot of bridges with Longhorn, and the OEMs might finally start looking for alternatives.

As for usability, Microsoft built it's market share on 'good enough', no reason Linux shouldn't. Of course since Linux doesn't have to worry about the tech support nightmare of supporting all those innovative features there's no reason they can't be available. I just don't think it'd be a good idea for them to be added to say, Redhat Enterprise.

I want linux to succeed with the masses, not for the sake of ideals, but because I want to keep using it. I rely on cheap, compatible commodity hardware to run Linux. Microsoft has hinted they want that market to go away now. That worries me. If there are millions of Linux users like me, I won't have to worry. So I say take the pragmatic approach to replacing Microsoft Windows (and let's not kid ourselfs, with 95% market share there's no one else to replace). Give users the familiar MS-Windows copy they want, and I'll keep my KDE desktop skinned to look like MacOSX.

Re:What I want in 2004 . . .

[ajs318]

This is why I like Slackware's simple packaging system. But I am no longer a newbie.

RPMs are dire, unless they are the common ones included in the distro. On my Mandrake box, I seriously prefer tar.gz packages because they are just more reliable.

The problem is maintaining an accurate database of what depends on what. Debian have rigorous testing procedures, they test everything so many times that it get out-of-date.

Two ways out of it:

1. sell CD-ROMs containing collections of packages proven to work when installed on anything with a kernel, because every dependency would already be meetable from the CD contents.
2. set up an online package base which automatically reports any installation difficulties to a forum where other users can offer suggestions, and once a solution is found, update the dependencies for that package for the benefit of the next person.

Option 1 has the disadvantage that you have to sell many copies of the CD to break even out of it, because you have to sell it cheaply enough that people will buy it from you rather than anyone else. Option 2 has the disadvantage that if it's reporting users' config info, it looks a bit like spyware {but from the practical standpoint, you can't just go to the vet and say "my dog is poorly", not take the dog with you, not describe its symptoms, and still expect a cure. But if you s/dog/PC/ and s/vet/phone support/ it's a different matter in most people's eyes}.

Re:The Future Fair...

[binary paladin]

"Linux will finally have a sound system that works and without it being a pain to deal with."

Kernel 2.6 has ALSA built in and with auto module loading it is really nice (well, nicer than it is without auto loading... which is still nice).

I really think that 2.6 is going to do a lot of good for the world of Linux. I've been running the betas and am VERY impressed with the responsiveness of my machine as compared to 2.4 and it's rock solid too.

It'll be nice when my distro runs it stock and the newer driver models are conformed to. When 2.6 is mainstream I think we're going to see a lot of good things for the Linux world all around.

Re:The Desktop Is Not Important Right Now

[Anonymous Coward]

I hardly think there's some linux "pie" where part of the resources are dedicated to servers and the other part is dedicated to desktop. If the linux community cannot motivate both parties simultaneously then that is its most serious flaw. The desktop and server users are basically separate anyway. You don't see a kernel hacker stop hacking and say, "Ok, now I've got to work on Gnome widges for an hour." No, he says, "Screw them. Gnome sucks. I just use fluxbox." That's how Linux works, if you want something done you tell everybody and hope somebody comes along that is capable of doing it and has the free time and generosity to do it. Given enough people this usually happens. Linux needs to keep finding jobs for people that haven't found jobs yet. It needs to keep its developers busy thinking up new ideas and new directions for them to work on.

That is how innovation works.

Re:The Desktop Is Not Important Right Now

[fzammett]

You've just effectively illustrated why it's unlikely Linux will EVER dethrone Windows on the desktop (yes, I said it!)... To use a basketball analogy...

If I have Larry Bird's championship Celtics teams on one side, and on the other I request anyone that wants to play show up... Even if Kobe Bryant, Tim Duncan, Jason Kidd, Karl Malone and Tray McGrady show up, chances are that Bird's team is going to win because they are the well-organized, focused and experienced TEAM. They know how to work together effectively and they have leadership who are focusing them on one or two definitive goals. The pick-up team has the better talent almost everyone would agree, and they all love to play, but if your not focused and a real team, your not going to win.

What Linux needs to have a chance is one or two guys that can rally the troops and drive the desktop effort in a huge way. The community needs to focus on Microsoft 100% to dethrone them. Hoping that someone talented enough comes along that can write pieces to compete with Windows XP isn't going to work. Someone needs to find those people, motivate them and get the most of their talents. Microsoft does this day in and day out. That's why they are on top as far as the desktop goes, and why they will almost certainly remain there for many years to come.

Re:My Bet Is On 2006

[Celvin]

I think we will see a lot more adoption abroad.

I assume you mean abroad as in not US. The comparison under is not quite fair, favors small countries and is in no way accurate, but who cares... My point is, Linux is widely adopted outside the US.

From counter.li.org:

USA: 94.90 users/million
Some states:
New York: 78.35 users/million
Texas: 90.95 users/million
California: 101.52 users/million
Arizona: 111.53 users/million
Maryland: 117.78 users/million
Pensylvania: 166.60 users/million

Some other countries:
Germany: 135.25 users/million
Spain: 144.31 users/million
Poland: 183.89 users/million
Holand: 215.54 users/million
Norway: 445.37 users/million
Estonia: 511.26 users/million
Finland: 610.85 users/million

My teacher was right, statistics can be fun, in a dry, inacurate sort of way... :)

Re:The Desktop Is Not Important Right Now

[Jason Earl]

Of course, in real life there isn't an arbitrary 5 man limit to your team. The championship Celtics were good, but they couldn't take on the best of the NBA greats all at the same time. Heck, they couldn't even compete with high school players if the opposing team had 500 guys on the court at the same time.

Microsoft has gotten to the point where they are competing with a huge percentage of their Windows developers. What's worse, if you come up with software that works well on Windows and starts making you a profit there is a good chance that Microsoft will copy your functionality and drive you out of business. This is why the commercial software industry is shifting towards Linux, that's the only way that they could possibly compete in the long run.

No matter how many sharp people Microsoft has working for them, if they turn this into a Microsoft-against-the-world battle they will lose.

Linux will win the desktop war in the same way that it is winning the server war. The key is to offer a product that is "good enough" at a lower price. Linux doesn't have to be better than Windows, it simply has to be "good enough."