Slashdot Mirror
Linux will have 20% desktop market share by 2008?
unmadindu writes "Siemens Business Systems, after conducting an extensive survey on non technical workers ("secretaries and managers, not IT people") is predicting that the Linux desktop will capture 20% of the market for desktop computers in large enterprises within the next 5 years. Senior program manager Duncan McNutt, who has overseen Siemens's testing of Linux desktops with users and administrators in enterprise settings, believes that the Ximian desktop and application suite, running on either SuSE or Red Hat, requires two days of training, which is the same as what most enterprises budget for a Windows/MS Office version upgrade. Interestingly, they used Ximian Desktop, instead of KDE, because Gnome, particularly Ximian's version, was "different enough" to set user expectations that the experience would be less like Windows. "
right now I have openbox3 with customized gnome-panel open, a transparent aterm and firebird with 4 virtual desktops open, and I tell you, it look prettier and works faster than any other system. especially now with the preempt patches to the 2.6 kernel and the new 2.4 gnome, all linux needs is games.
I would love to see it, though I think it depends more on what MS is capable of delivering with Longhorn that what Linux can do. My guess is that if the economy is still in the crapper, and people are still using a decidedly client server computing model, then upgrades to a new MS OS are going to be slow on the uptake. We need a paradigm shift in IT, something new and wonderful needs to happen. Linux desktops should be going for new and wonderful, not same old same old.
All these so-called 'predictions' are useless. No-one can look into the future and especially in the fast moving world of hard- and software the Next Best Thing is always just around the corner, so why do people take the time even to read predictions like this?
-- Cheers!
They used Ximian Desktop because the menu interface is ordered with a more clear naming than KDE.
My 0.0002 euros
While the article is a bit thin on details on this, I'd be curious to know what this extends to. Is it just the look of the widgets? Questions like single vs. double click? Menu layouts of the standard applications? Did anyone make this experience before when trying to convert folks to Linux?
It's a shame that we don't have results of a survey like this from before and after the SCO storm hit. It would probably very useful when it came time to extract some damages from the pump and dump crew.
I for one am scared that the long term effect of the SCO lawsuit will be a slowing or reversal of linux's creep towards the desktop where the final battle with closed source development will be.
PornStarGuru
Oh come on indeed, one thing is certain... you need to pay heed to past false predicitons, and at least honor current trendlines.
Google HTTP shows Mac ALWAYS 9x larger tahn Linux every year since 1995.
therefore for linux to be larger than its relationship versus mac i find laughable these predictions.
In 1996 linux was supposed to finally overtake mac.
In 1997 linux was supposed to finally overtake mac.
In 1998 linux was supposed to finally overtake mac.
In 1999 linux was supposed to finally overtake mac.
In 2000 linux was supposed to finally overtake mac.
In 2001 linux was supposed to finally overtake mac.
In 2002 linux was supposed to finally overtake mac.
In 2003 linux was supposed to finally overtake mac.
Now these hilarious predicitons of 20% desktop market when Google OS in http referrals (OS never spoofed usually) shows that linux still never got close to mac.
That is why tens of THOUSANDS of shrinkwrapped commercial mac products exist, while almos NO commercial shrinkwrapped Linux products exist.
Marketshare.
Mac users do buy abnormally large amounts of software per machine sold, but even taking that into account, the lack of commerical Linux shrinkwrapped software should indicate SOMETHING.
Until Google shows linux gaining, i will chuckle at these predictions.
If they said in 5 years everyone would be using FreeBSD-Mach-Darwin OS (Apple OS X) that might be believable... but not 20% desktop Linux.
In addition to this statement from Siemens, I wonder if there is any company that has ever evaluated the time lost in desktop use using Windows 98/2000 on PCs in an enterprise-wide level compared to Linux, in a typical day's work, and that which is lost with linux. To be fair, this comparison ought to be with controlled environment (well set-up systems, users are only Power Users and therefore unable to install applications themselves, etc..).
This would result in something like:
Setup: Intel 500MHz/1GHz Desktop (or laptop)
Cold Boot Up
Login time
starting Lotus Notes/Outlook (viewing emails/starting new messages in Notes is historically long!)
opening word processor 1st time/next time
opening spreadsheet first time/next time
opening presentation tool first time/next time
opening web browser first time/next time
shutting down
rebooting (yes, even in linux this may happen!)
number of rebooting
etc... (applications in Enterprise environment, not home use, hence no video viewer or filesharing software for example. IM is not yet a universally accepted tool in my experience either)
If workers in a 1000-employee company were asked to monitor all these tasks for a whole week, half of them on linux, half of them on Windows, this should return an average that's actually measurable and would start making sense.
Does this exist anywhere?
There is no god
I think it's a save prediction that in 2008, the state of the Hurd will be "production ready in about half a year". But still, you are right, this isn't about Linux, it's about Gnome and KDE, whatever OS they are running on.
Programming can be fun again. Film at 11.
This will NEVER happen by 2008.
As another reader put it "Oh Come On".
Even if linux were to get to 10%, MS would release a new stripped down version of windows and office for a reduced price to cut into the market that this study says is going to flock to linux because it only takes 2 days of training.
What happens when these people get sent a MS Project file and can't open it, or what happens when they call the support desk and the person tells them to open their c:\winnt folder??
Come on people, you are starter than these posts.
I wounder if that desktop expansion will be more at the expense of Microsoft or Sun (and to a lesser extent SGI, I suppose). Replacing relatively expensive Solairs desktops with Linux is straightforward; replacing M$ generally requires a shift in how applications are delivered (e.g. a move to web-based or Java applications).
Why does it have to be installed in large scale environnments for productivity gains? The article states that the training required is the same. If that is the case then it should be good for any size business???
Training is only part of the cost structure for any IT deployment. The cost savings of desktop Linux are due primarily to it's UNIX heritage: its security model, centralized authentication, network filesystems (both NFS and AFS), and it's inherent ability to scale from thin client to full workstation without any back-end changes to user accounts. This is all traditional 'NIX stuff going back to late '80s early '90s Workstation fare.
Why this matters is that an organization doesn't see significant cost savings along these lines until they hit a threshold deployment size, nor are the savings linear from the bottom up. Ten Linux ('NIX) workstations don't save the same percent of money in an IT budget as do one hundred. One Hundred saves less as a percentage as one thousand. I don't have numbers, but I've seen the savings first hand - the bigger your deployment gets the greater your savings due to reduced overhead (IT staff) costs.
This is why I don't think we'll see Linux take off as a desktop platform for most small businesses, but we will see it deployed throughout government and large industry players. It will likely move from foreign markets to the US as well, simply because third world industry is under heaver cost constraints compared to the US. But like all network effects, as industry uses it abroad, US players will have to follow in order to maintain some level of compatibility' most likely we'll see US players install OpenOffice and then it will mushroom from there.
JMO.
Cheers,
--Maynard
We all might be running Sconix by then.
One way or another those jokers are done. Nobody is going to do business with people who consider contracts weapons any longer than they have to. Assuming SCO does their job and kills Linux, Sun and MS will finish them off. SCO would just turn on them next. Do business with SCO and you'll get sued. Everybody knows it.
Agreed lots of things could happen. I don't think SCO's survival is going to be one of them.
How to gain real marketshare for Linux on the desktop.
Standardize all hardware installation and removal in one place across all distros.
Name changes that non-it people get. Grep makes sense to IT types, but few outside IT are going to know what it means. Similiarly, I shouldn't have to explain that eth0 refers to their Network card and so on.
Improve Wine. You can give me a hundred stories about how with your uber-133t skills you get a certain archaic package to work under a certain distro and that lusers don't need graphics anyways. This is exactly the type of attitude that will keep Linux from the masses. They want to be able to use their programs, and most could care less what OS their using (how many times have you talked to someone who didn't even know which OS they had?). If they can happily use the same programs they used before, they could well not even notice the OS.
Most importantly of all, all versions of MS office must work seamlessly. This is the standard in the business world, and StarOffice, OpenOffice are poor substitutes. They don't want to learn the quirks of these packages, they just want to use MS Office. Nothing is more important for gaining marketshare than this.
Drop the attitude. The attitude that many newbies encounter is more than enough to send them back into bill's not-so loving arms. When someone is trying Linux they far too often run into someone who an elitist that thinks they should not only know *nix inside out, and be a programmer to boot. When joe-sixpack gets told to go RTFM after asking what a tarball is, he's going to get indignant and goes back to what he knows - windows.
Have a resource available to those who come from the Windows world that tells people in plain English what the Linux terminology is for equivalent ms / windows functions. Also have this resource list programs like gimp that can replace their old windows programs. A frequent complaint of those that try switching to Linux is that they can't do what they used to freely do under Windows. Slashdot types will respond, of course they can, they don't know what to use. Well, how would they know what to use?
The subject of this reply sounds like a troll, but considering this fact it might actually be sooner. All chinese civilians will probably be 'encouraged' to run chinese s/w as well. With 10^9 inhabitants and a growing market for personal computers, China may make a bigger dent in the statistics than Microsoft would like.
Wenn ist das Nunstueck git und Slotermeyer? Ja! Beiherhund das Oder die Flipperwaldt gersput.
Interesting point. The differences may be just as important to user acceptance as the similarities. Reflects a point I've tried to make in management discussions: Linux is not better now because it's like Windows, Linux is better because it offers advantages over Windows on many levels. So far I've been the token open source advocate, but the interest level is definitely on the increase. It's not lost on the boss that when the virus-o-d-day comes around our RedHat servers stay online.
Still some acceptance hurdles to cross and some technical improvements needed, but we're getting there. Amazing to me how fast it's gaining ground.
Viva la Penguinista!
That's our life, the big wheel of shit. - The Fat Man, Blue Tango Salvage
Furthermore if Linux holds 20% you're going to have compatibility problems up the wazoo(sp?) The reason everyone uses Microsoft products is because it works[sic] so well together.
Look at the Linux desktop four years ago when I migrated my parents to Linux with RedHat 6.1...... And no, they are not great with computers.....
Look at the Linux desktop today where finally RedHat, SuSE, et. al. are trying to push for a Linux desktop market. This would have been unheard of 4 years ago.
We already have early adopters in Muenchen, and other places, but the large-scale deployments seem to generally be governments, while the small scale deployments tend to be smaller businesses. Here are the pro's and cons to Linux vs Windows:
Cons:
======
Lots of software available out of the box for Windows. Many well developed desktop applications.
Microsoft RAD environments have a much larger mindshare than their UNIX equivalents (TCL, Perl with GTK, etc.).
Company may have large number of legacy VB applications, and I have still had serious difficulty getting many Win32 applications to install or run on Linux using WINE.
Pro:
==========
Flexibility: This is open source's killer app, IMO. With Linux, you can download a set of ISO images from the net, roll out a pilot program on existing hardware without having to procure anything. The same holds true for the BSD's as well. Of course if you don't buy it you get no support, but you may not want to pay for the support for a 3 month pilot program since you can then buy things later when you have figured out exactly what you need.
Linux is far more admin friendly than Windows is because the "toolkit" approach that Microsoft disparages gives far more flexibility than Microsoft's "End to end solution" approach. Again, this comes down to additional flexibility for a business who can now easily use existing solutions in new environments.
In essence all the "Software for an Agile Business" ads aside, it is clear that open source gives businesses MORE agility and flexibility than any proprietary solutions, and that this is most pronounced when compared to Windows.
LedgerSMB: Open source Accounting/ERP
One of my wife's friends wants me to install Linux on her machine just so she can play Frozen Bubble. They are all addicted to that game.
I started using Linux for coding about a year and a half ago, and switched to using it as my main OS a year ago.
I had various problems getting odd bits of hardware to work, etc, etc... nothing too serious. I found people were generally helpful. The one offputting thing that happened was this:
I use a chat program that I wrote under Delphi in Windows... making it pretty much impossible to port. Under Windows it binds to port 23 to let people connect to it with telnet... obviously impossible under Linux, but I didn't know that at the time :-)
Anyway. I asked how I could get around this -- and the person on IRC said, well, you could start it as root and drop priveleges. And then I said I wrote it and was running it under Wine... and the response was one of disbelief. Why I want to run anything under Wine?... it was 100% necessary to my switch to Linux, but they weren't interested.
So -- switching operating systems isn't easy, particular for a home user. The fact that I'm a compsci student, plus the fact that I was persistent, mean I made the switch okay. Much happier with what my computer can do now :-)... but you're absolutely right about the attitude of many current users. Relying on Windows doesn't mean you're stupid or weak; it's a tool like any other.
If something should go wrong under the hood, like the internet connection drops, God help her if I'm not around. And she could not have set the system up herself. But with large organizations like the article discusses, that's not the end-user's problem.
Don't drop the soap, Tommy!
That's probably a gross understatement. When Linux breaks the desktop (like it did for servers), a couple things will happen.
First, development of the desktop will be relatively much cheaper, due to the large mass of users. More slickness and more applications (finally...)
And second, MS will be in deep trouble. They can't keep hiking the price to sustain their profit level. They'll also have trouble reducing the prices significantly, as shareholders would panic. And if they introduce a new desktop OS, it'll have to compete head on with a seriously tough enemy - the perfected (and still free) Linux desktop.
While Linux lives just fine with 5%, 10% or 20% of the desktop, Windows doesn't - a major drop MS-Windows marketshare would cause the confidence in the platform to erode, and thus create an interesting snowball effect, leading to great savings and great freedom :)
Optimist, and proud of it!
I'm in a Unix state of mind.
"is predicting that the Linux desktop will capture 20% of the market for desktop computers in large enterprises within the next 5 years."
Ok, but what is the size of "large enterprises"? This could still amount to a small portion of the entire world's desktops.
In 2002 Linux did overtake mac.
;) For the most part, I agree.
Linux is part of the free software movement, Linux users are disproportionately hostile to shrinkwrapped commercial packages; further Linux distributions come with tons and tons of software and you can download software for free for just about everything.
Adding to your comment: You can't buy Photoshop for Linux, but you can install GIMP free for Linux. (ok, won't be doing any CMYK work) You can't buy Norton AntiVirus for Linux yet (although eventually, it will be needed even if Norton has to write the viruses themselves
So few commercial packages are purchased for Linux, because so few packages are available. The real key is when they start porting over business software for Linux. If Peachtree had a Linux port, we would have it in ONE day. ACT is another example. And yes, I know, there are free alternative, but for most of us, the alternatives are not alternative.
MOST applications do have alternatives in Linux, and many are better. Its just these larger applications that need to be developed. Ironically, it would appear that if you DID port Peachtree for Linux, it would be then be trivial to port to OSX.
20% in large enterprises by 08? IF the big apps catch up, I can see this. If Open Office can keep pace. If Gimp makes another leap. The smaller applications are already there, and it already is easier to maintain Linux if you know how.
Tequila: It's not just for breakfast anymore!
Seems like bloody slow progress to me, given that Linux has been around in a usable form since 1993 or so and that in 2008 it will be about 17 years old. Linux had a usable desktop in 1994, or at least a lot more usable than Windows back then. Still I suppose this process is social rather than technical.
-- Ed Avis ed@membled.com
The only hope is to have a version of WINE that is so reliable and capable that it not only will run everything, but is also usable and fast enough that it develops a solid reputation for being "the" answer to this problem. Anything short of that and businesses will not convert to Linux--why should they go through the pain and expense of converting to an OS that doesn't meet their needs, even if has a zero marginal cost and superior reliability and security?