New Numbers on Linux Market Share Soon

prostoalex writes "New numbers on Linux market share are due this week. As far as global PC market is concerned, Gartner claims 5% of all PCs shipped this year ran Linux OS, although by the time the PCs were actually on the user's desk, only 2% of them run Linux. In the server world IDC estimates that Linux-powered servers comprise 28.3% of all server sales in 2004."

Comments

The other way around?

[BladeMelbourne]

Gartner claims 5% of all PCs shipped this year ran Linux OS, although by the time the PCs were actually on the user's desk, only 2% of them run Linux.

And how many people buy PCs with Windows on them, and immediately format the disk(s) and install Linux?

Re:The other way around?

[nmoog]

I dont know about that. My computer was assembled by me and contained no OS when I bought it. Probably a good chunk of linux users are in the same boat.

And if you consider the stats from Google's zietgeist, it really can't be too many.

Re:The other way around?

[jmv]

That's especially true for laptops -- try to find a laptop that ships with Linux. There are some, but there are so few they're not likely to have the config you want.

Re:The other way around?

[shaitand]

Pirated copies of windows amount to a pretty much insignificant amount of the desktop.

The reasons are simple, most users aren't capable of installing and OS period. Most of the "computer savy" kids aren't able to install an OS either, they are considered savy because they can download and install Kazaa, download some music and burn it to a cd. This convinces their parents and family they are a computer whiz.

The other reason is that windows typically comes on a new computer. When you couple this with number one, you quickly realize that the knowledge required to find the data and transfer it from all the various applications one might run is far more complicated than installing the OS alone.

Remember, those who do have knowledge to do these things are typically techs and admins. At work they won't be pirating windows. Outside of work THEY might pirate windows for themselves, friends and family, and might mostly associate with techs, their friends, and family. This means that your typical tech/admin will see lots of pirated windows copies floating around.

Remember also that all the techs, all their friends, and their families probably add up to less than 1% of the total desktop market.

Wasn't this the opposite argument we were making?

[vondo]

Isn't this the opposite of the argument that Linux users have always made. That because it is so hard to get a PC (as opposed to parts) without Windows, that the number of linux installs running was higher than the shipments?

I mean, really, what evidence do they have that hordes of people are buying machines with Linux pre-installed just to go through the pain of installing XP in order to save, what, $40?

Granted, a lot of machines shipped with Linux aren't running the version of Linux they shipped with, but I find their statement hard to believe.

Re:Ship % should underestimate, not overestimate..

[Pieroxy]

And here goes the danger of thinking that your family if somewhat representative. Or the slashdot community for that matter.

NX Bit?!??

[Teancom]

What the hell? As long as Linux doesn't support NX, which *will be* supported in the *upcoming* SP2, it will be seen as deficient in comparision? Okay, ignore the fact that they themselves state there are already patches for Redhat, while SP2 hasn't actually shipped. I'm just wondering what the fascination with NX is. I mean, it's a nifty idea, but I can't imagine anyone getting down the wire of choosing between XP or Linux as the right tool for a job, and deciding on XP because of NX. I mean, come-on. This is just idiocy. Not that I had any respect for Gartner to lose, but if I did, there it went... There are plenty of differences, strengths and weaknessess on both sides, to differentiate between XP and Linux. Supporting the NX bit is not one of them at this time.

Re:NX Bit?!??

[bigberk]

You named one component. ActiveX is a subset of the IE component. Other than IE, do you have any other insecure components to mention?

Yes, I can name other insecure components. How about USER32/GDI32 and its messaging design that allows privilege escalation/leaking, demonstrated by the very real Shatter Attack. This demonstrates fundamental flaws in the Win32 messaging design, which can not be fixed without radical changes to Windows.

Note that I'm not arguing by talking about insecure/vulnerable libraries, like MS crypto or RPC which has led to dozens of major exploits. Equivalent vulnerabilities in portmapper or OpenSSL would lead to similar problems on UNIX. Rather I'm talking about fundamental design flaws that allow privilege leakage -- (1) IE integration (a huge one), (2) Win32 API messaging, ... any more?

Re:Ship % should underestimate, not overestimate..

[Entropius]

Well, yeah. I know that. But I'd think the number of people adding linux would be greater than the number who buy linux boxes and format them, just because there are so few linux boxes sold.

Re:Ship % should underestimate, not overestimate..

[Kenja]

Yea, because no one is going to buy a low cost Linux computer at Walmart and slap a pirate version of Windows on it. Nope. Never gona happen.

Re:WindowsXP is free...

[nwbvt]

Legal issues aside, most of us do not use Linux because it is cheaper, we use it because it is a superior operating system. Hell I already have the Windows OS that came with my PC (so it might as well be free for me as well). The fallacy that people only use Linux because they are cheap unemployed hackers is just that, a fallacy.

Re:Ship % should underestimate, not overestimate..

[Xabraxas]

...and no one is ever going to buy a windows computer and slap linux on it...nope, never going to happen.

Re:Gartner?

[Ami Ganguli]

They're a pet peeve of mine. It seems like what they do is interview CIO types about their opinions on various technologies and then turn that information into speculation about where the industry is going.

The result is a bunch of very credible sounding propoganda that reflects all the biases prevalent among their target audience: CIOs who need backing for their opinions. The CIOs naturally buy the reports and use them to pursuade other people in the company that the CIO's favourite pet project or technology is "industry best practices".

Gartner reports tell a lot about what people who worked in technology ten years ago (and have since moved to management) think. They consistently overlook trends that are bubbling under the surface, obvious practitioners, but not yet noticed by management.

If you want to know what your boss thinks about the industry, read Gartner. If you want to know about what's really happening, read the Usenet group that deals with the specific technology you're interested in.

Re:Ship % should underestimate, not overestimate..

[golgotha007]

I'd think that the percentage of computers that actually run Linux would be higher...

not only that, but how the heck can they possibly know how many people are running linux or not running it?

I understand they can count the number of linux pc's going out the door, but how do they know people are installing winxp or not?

also, how do they know how many linux desktops are out there? are they just counting how many copies of mandrake were bought at CompUSA?