Gartner Claims Less Linux Than IDC

Quite a few readers submitted news about a recent Gartner study which implies a far-lower-than-usually-supposed percentage of servers running GNU/Linux. Reader mfarver, for instance, writes: "Only 8.6 percent of servers shipped in 3rd quarter 2000 were running Linux, claims a recent Gartner Dataquest report. A previous study published by IDC estimated linux held about 24% of the server market share. Unsurprisingly the Gartner study was partially commissioned by Microsoft." Roblimo has penned an interesting piece up at NewsForge about why those numbers might smell a little funny. Hint: how many machines have you bought running any Free operating system from the get-go?

Comments

Shipped 'were' only includes pre-installed?

[Anonymous Coward]

The article states that the adminstrators surveyed were asked what OS is currently running on the servers they purchased over the last three months, not just what shipped with pre-installed OS.

OTOH, much of the numbers rely on their definition of 'server' as hardware that was built as a server-grade system, not just any old PC pressed into service.

A company that is likely to cut corners and use Linux instead of a commercial OS is also likely to cut corners and use a commodity non-server-specific machine rather than buy a 'server-grade' hardware platform.

Re:Shipped 'were' only includes pre-installed?

[Twisted Mind]

Yes, I was also wondering.
In my company we have one Compaq NT-server as file/domain-server (unfortunately ran by some extrernal admin-company that refused to administor a machine remote), and one (and over some time two) linux machine installed on desktop-machines serving as mail/firewall/proxy/intranet-server.

The first NT-server is essential but the linux-machines just as well.

Re:the real world

[Anonymous Coward]

All I do anymore is *replace* NT with BSD or Linux

Is that why you can't keep a job?

geekforhire

Perhaps you should spend less time fucking over some poor company's network and more time with your Playstation.

TRUTH

[Anonymous Coward]

The fact is that no company I've worked with has purchased their Linux hardware. The whole idea behind using linux was that we could get more bang out of our existing systems. Turning NT 4.0 servers that crashed weekly into systems that never needed to be rebooted. Only one company ever tried to buy a system from ASL and I quickly stopped them. Pointed out that we could simply upgrade some existing rack hardware that was collecting dust in the storage closet. I don't like the idea of some other tech setting up a box I admin. Its a matter of trust, and I trust no one with what could make my job easier or a nightmare. Here I sit in my cube, on my Comcrap workstation that I wiped out W2K on, sitting next to my Ultra 10 running (yes again) Linux. All of our servers run it, other then the ultrasparc server, and that two would be _upgraded_ to Linux if I could take it down. Today the only W2K systems we have are workstations and the lobby portals. Due to the fact that most of our workers had been duped into Windows a long time ago and I swear sometimes it seems it would be easier to get them off crack. That and the portals require it due to a lack of macromedia director compatibility. In short we, the techs who drove the industry now work to drive it elsewhere. We where lied to. This isn't something we can forgive or ever forget. It took many years to build you up Bill, now watch how long its going to take us to tear you back down. Its only a matter of time. So steal that new idea, crush that new competition. "Innovate" as you have always done...

Re:TRUTH

[Lurker]

Actually, where I work, we've bought 4 Dell 2450's and 2 2550's preloaded with Linux so far. Granted, I still rebuild it from scratch... Aside from preferring to install it myself, they also won't set it up as a Raid10 from the factory.

You can't trust Dell to do what you ask in any case. Where I work we're about half-and-half MacOS and Windows with our lone unix server (MacOS X Server) handling most of our intranet server needs (we're not a big shop.) I order all our Windows machines from Dell. When we were using NT, I would always specify "Windows NT 4.0 Service Pack 5 on NTFS". What did I always get? A two-partition setup: partition 1, with the OS on it, was always FAT and partition 2, which usually had some apps like office, was NTFS. A lot of good this does me. The first time this happened I actually (naive, oh so naive) called Dell and told them they had obviously done it wrong. I explained that what I expected to get was one partition using NTFS that spanned the entire disk. They responded, of course, that they couldn't do that, as NT 4.0 SP1 (which is what the boot CDROM that came with the computer was) couldn't make NTFS partitions larger than 2GB, you needed to have NT 4.0 SP3 or higher installed to do that. No matter, I always reformat & do it myself anyway.

Now that I order new Windows machines with Win2000 Pro on them, always specifying "Windows 2000 Pro on NTFS", expecting that we would get a machine with a single NTFS partition, what do we get? FAT32, thank you very much. Oh well, I always reformat anyway, to get rid of the extra cruft Dell installs.

Re:TRUTH

[naChoZ]

Actually, where I work, we've bought 4 Dell 2450's and 2 2550's preloaded with Linux so far. Granted, I still rebuild it from scratch... Aside from preferring to install it myself, they also won't set it up as a Raid10 from the factory.

Re:TRUTH

[naChoZ]

Yeah, whaddya expect. You know it's some flunky running ghost or something, stuffing images on drives all day long.

One of the reasons we stick with Dell though is because of the 4-hour onsite guarantee you can purchase. Otherwise we'd just go with penguin computing since the prices are still competetive and they'll ship with a raid10 config already set up. Plus there was even a nifty little thing on the web page to allow you to customize your file system layout.

Re:Bullshit.

[Anonymous Coward]

You have to understand the service that Gartner provides. You tell your manager "We should switch to X for task Y". Your manager tells his boss "We're considering a switch to X". His boss thinks of his own boss and says "Can you justify it?". Well, that sounds like real work, so your boss looks up the Gartner report that indicates that X is a good idea for task Z, selectively quotes it and gets the project approved. You are then happy.

Left out our 32 servers running Linux

[Anonymous Coward]

I work in a university in the UK. We use linux as our file and print servers. We have 32 servers doing the job in total, and all of it was network installed (ftp and http) through the UK academic mirror site.

Some would argue that for such a large scale installation we should have vendor installed servers. But we found that we wanted a very customised solution, and without putting the vendor down, we felt that we had enough expertise within. At the end of the day we were in complete control.

I shouldn't be too worried about statistics, and I'm sure we weren't counted.

Re:Left out our 32 servers running Linux

[clare-ents]

We've got three linux servers here. They were all shipped with NT Server, but P100-200 machines are a bit slow for that now. They're only doing small stuff, a little webserving, RADIUS, quake serving, some log file processing and a few other automated tasks, but we only ever bought NT Server for them.

We're vaguely thinking about a support contract to cover the day when the knowledgable person is out the office [me], especially as they don't cost much compared to NT Server.

Re:Linux Counter?

[drwiii]

This guy sure did.

Re:Gartner smells like Ziff Davis

[The Man]

IBM had 90% market share, a giant marketing machine, and every PHB in the world. "Nobody gets fired for buying ___________" was first said about IBM. They had FAR more power than Microsoft ever has, and they fell on their heads. They've clawed their way back by giving people what they want.

Re:Gartner smells like Ziff Davis

[The Man]

I won't address much of the standard rhetoric in your post, but this really stood out:

If [Gartner] produces poor quality research on a regular basis, people will stop subscribing or purchasing its reports.

Logically, in a marketplace of informed, intelligent consumers, this makes perfect sense. Unfortunately, substitute "Microsoft" for "Gartner" and "software" for "research" and suddenly the apparently obvious, rational statement becomes a joke. I'm sure we can all name dozens more examples where obviously inferior products nevertheless generate substantial business.

I don't know whether Gartner's study or IDG's or Netcraft's or all or none are correct. And I really don't care. I don't choose my products based on market share. After all, isn't the whole point of Gartner that people will choose their research because they're well-known? If you apply the same standard to operating systems, what does that say about you?

Don't fret the small shit. Focus on quality. Every business that's successful over a long period of time worries about quality first. A company can succeed by raping customers for a limited length of time. Eventually it has to produce a product people actually want, or go away. IBM knows this all too well. Microsoft will learn it too. So move along, please; there's nothing to see here.

Re:Netcraft Statistics

[The Man]

"Linux ruleZ!" "No way, d00d, Linux suX0rZ!"

Dear God, man. It's ok to point out that someone's wrong without being such an ass. Sure Apache runs on lots of different platforms - in fact, almost every major platform there is. It's also true that a large proportion of the sites using it are doing so on Linux. You probably figure they've been hyped into obvilion or that the low cost was attractive. The other poster probably just thinks that Linux is 31337.

So what? Linux is a very nice OS. It's unfortunate that it's mostly used on a decisively inferior hardware platform that effectively limits scalability, performance, and reliability to a few times that of the best Microsoft products. Can you do better than Linux on a peecee? Of course. Can you do better than $FOO on $BAR? Of course. You just have to spend more money. I've used all those "actual enterprise OSs" and I've adminned them, deployed them, and helped people make money with them. Frankly, I'm not real impressed. The only Unix I'd ever recommend over Linux is Irix, simply because nothing else will run on the beautiful purple monsters SGI makes. If you don't need those monsters, you can do just fine with a Sun running Linux. Choose the hardware you'd want if you were to run Solaris, then buy two models down for equal performance.

But I digress. Use what you like, but don't insult my choice of systems just because it's the same as that of someone who can't justify his.

Leave the PHBs to do their thing

[The Man]

There was no marketing of Free Software, nor Linux, and surely not "Open Source" before 1998 or so. Nada. Zilch. No multibillion-dollar IPOs. No bandwagony press releases from IBM, SGI, HP, or anyone. No bandwagon at all to speak of. And yet in 1997 Free Software was alive and kicking, doing well in spite of adversity and ignorance. Linux-based operating systems were beginning to take off. The FSF was as alive as ever, kicking out great amounts of high-quality software for all to use and enjoy. The various BSD projects were around, too, producing very much the same product they are today, and not of much less quality.

What's the point? The point is that this community was already a vibrant, rapidly expanding one well before anyone decided it was the Next Big Thing. Long before you could read about it in the New York Times. Long before Slashdot became what is is now. Long, long before would-be nerds pondered the latest market figures for Linux-based systems from the Gartner Group and IDG.

Let the PHBs be. The community will survive and prosper with or without them. If everyone who only uses Linux or Free Software altogether because of advertising, press releases, support of large corporations, or mainstream media attention stopped using it right now, what would we lose? Well, a few fine folks would lose their jobs. At least one or two more "Linux vendors" would probably close up shop - but that's likely to happen anyway as part of a much broader downturn. Our market, if we cared to measure it, might shrink by 30% or 40% or even 70%. The software would not vanish from FTP sites - at least, not from all of them. The sources would not go away or suddenly become closed. And if I had to wager, I'd suggest many of the people no longer paid to hack this stuff would still spend time on it. In short, things would - and will - go on as they always have. The bubble was just that. You'll see it years from now as a spike in the plot of market share with respect to time. But don't worry the spikes or the valleys, watch the long-term trend. That trend has been toward a very high-quality set of products enjoyed by more people for a long time.

Forget Gartner, forget IDG, and forget the silly people who think they matter. Step back and look at what you're running today, and enjoy the fact that, all the negative press notwithstanding, you've got Free licenses that never expire for some damn fine software. Relax, and go outside for a breath of fresh air, lest the flames ruin your good mood.

Re:Study automaticly bies

[The Man]

Linux usefulness starts to fall off at a certen point on the high end so sales of extreamly high end is likely to be Windows NT only.

It is true that the usefulness of Linux diminishes at the very high end. But in the realm we're talking about, nobody even thinks about NT. In fact, NT's usefulness is minimal (even by comparison with some NT "base") in midrange servers and nonexistent at the high end. The hardware NT (and x86 Linux) support tops out in the lower midrange of servers. A maximal 8-CPU peecee is a minnow in a school of sharks. At least Linux gives you an option to move beyond the limits of the peecee - an option NT once offered but no longer does.

No, toward the high end you see things like Irix and AIX and MVS and OS/400. Definitely not NT. But that doesn't mean the numbers aren't meaningless; they most certainly are.

Re:Think about this...

[The Man]

Ahh, this old thing... This would be a simple matter if all admins were of equal value. But that one experienced Unix admin can run a large department with nothing but cron and a tape monkey. And when something breaks, he'll solder together a magic doodad to make it all work again. The 5 monkey admins can read Microsoft's web site all day and all night, but there will always be serious limits on what they can really do. Some limits imposed by the OS itself, many more by their own lack of experience.

I don't know about you, but I would find it very difficult if not impossible to pick up meaningful experience in a world in which there is little relation between action and consequence. Two people can do the same thing on two identical Microsoft systems and get completely different results. More interestingly, you can do the same thing twice on the same box and get different results. Most maddening of all, you can do something, then reverse it, and end up in a different spot than the one at which you started. This kind of environment is not conducive to the type of learning that really good admins go through. There's actually a nicely similar situation discussed in Carl Sagan's Cosmos - he argues that if there were no rules of physics, science would be impossible because we could never learn by experimentation, and no experiment could ever be duplicated.

This entire idea, in fact, also skirts very closely the Monkey Rule. Specifically, a monkey can never be a good systems administrator. Sure, he can handle the common cases and push the right buttons if he's well enough trained, but the moment something out of the ordinary happens it's all over. At the risk of sounding disrespectful, I'd suggest that your 16-year-old sister most likely suffers from the exact same problem. So, no, I wouldn't likely pay her my salary for doing my job, and I'd hope no manager would either. It's not just that you need more admins to manage a dodgy infrastructure. It's back to Brooks: 5 people are not always better than one. Especially if those 5 people don't or can't contribute anything to the solution. Are there managers who've never read Brooks? Surely. They don't last long.

Yep.

[hawk]

They found that only 8% were using GNU/Linux. question answered. Never mind the 17% that were using Linux . . . .

:)

hawk, wondering what good a system with nothing but the linux kernel and ghe GNU utilities would be, anyway . . .

The basis of statistics...

[sheldon]

It appears from my reading that the Gartner numbers are based upon going out and surveying a variety of IT companies ranging from small to large. They asked them how many servers did they install and what OS are they running. Given the nature of Linux distribution, that is really the only way you'll ever know who is using what.

Whereas the IDC numbers were "extrapolated" with some pretty wild assumptions. I recall reading that they just automatically made the assumption that every copy being purchased or downloaded was installed to 15 machines. How can you know that?

There is also an issue of what's the point of these numbers. Looking at marketshare tends to ignore installed base figures so I assume that the point is to show what is growing, where the market is moving... i.e. where the money is.

As such, I think Gartner raises a very valid point that a Linux install that doesn't have some sort of support contract going along with it, says "what's the point?" Gartner points out that not many companies are going to do something like that anyway.

But more importantly, if the installs aren't being accompanied by support contracts, then they really don't play a part in accounting for "where the money is".

If the money isn't there, it's not useful for someone like RedHat to show market share growth. At least not to encourage investment, maybe to get warm fuzzies.

It's all relative. Honestly, from what I've read the Gartner numbers are much more supportable by facts. The IDC numbers appear to have a tint of "I pulled this out of my ass monday morning because we wanted a headline."

Re:The basis of statistics...

[nagora]

As such, I think Gartner raises a very valid point that a Linux install that doesn't have some sort of support contract going along with it, says "what's the point?" Gartner points out that not many companies are going to do something like that anyway.

But more importantly, if the installs aren't being accompanied by support contracts, then they really don't play a part in accounting for "where the money is".

On the other hand, a valid issue is "where the money isn't", ie, what money is MS missing out on. In that case the lack of support contracts is unimportant.

In general Linux gives the user much greater opportunity to do their own support, which is what my company does. This is a problem for RedHat as much as it is for MS but it's not a problem for those of us using Linux as their server platform. We've got the source code!

TWW

Re:Gartner smells like Ziff Davis

[sheldon]

Hmm, you must have read a different article than I did. The one I read indicated Gartner surveyed the IT depts to see what they were actually using.

My boss has a Gartner account, and I'll have to try to get access to the real study.

Re:Pre-installations were not involved in the numb

[sheldon]

I think that's because IDC made the claim that 25% of servers ship with Linux. Hewitt is questioning that. It's especially difficult because IDC tried to include downloads in it's "ship with" statistics.

Re:Pre-installations were not involved in the numb

[sheldon]

Claim #1. XP breaks MP3.
This is pure unadulterated FUD. Beta XP shipped with an MP3 encoder to test the ability to plug other audio encoders into Windows Media Player. The encoder they shipped was limited to 56kbps because that is the limit before you have to license the technology from Fraunhoffer. The latest XP builds do not include an MP3 encoder, and there was never any plans to ship with one. Windows Media has never had this ability, but they wanted to allow for third parties to extend the product in such a way.

#2. Breaks CD Burner Software.
I don't know the specifics. But given every Windows upgrade known to man has broken CD Burner software it does not surprise me. This is because CD Burner software still tries to operate at a low-level API. I don't know why they can't standardize on a device driver API, but Roxio, et al does seem to want to do that.

#3. Prevents you from upgrading your computer.
It's unclear at this point the details. I'll give you this point.

#4. Requires activation. Requires giving out info.
Point #1 is true, Point #2 is FUD. The activation does not require giving out personal information unless you choose to register your product at the same time.

#5. Is a very MINOR technical upgrade from Win2k.
Well this is true, but Windows XP is designed as an upgrade from Win95/98/Me, and in that case is a very MAJOR technical upgrade. Again this point is FUD.

#6. Has a GUI that would make anyone outside Miss Shirley's Romper Room throw up.
One of the benefits of growing up in Chicago was being able to see Romper Room and Bozo live. So perhaps I've been warped by those shows, but honestly having worked with the WinXP beta I think the GUI is pretty nice. Sorry, I'll have to call this FUD as well since it's obvious you haven't used it.

Well so far it seems the score for your post is:

Fear Uncertainty Doubt: 4
Insightful Opinion: 1
Pointing out the Obvious: 2

Not bad for one day of trolling.

Re:Think about this...

[sheldon]

"Lets face it, my 16 year old sister can learn to be a windows admin from one of those tech schools that'll make you certified in 3 months. "

Actually that's not true.

"And you can pay 5 of those newbie admins for the price of one unix admin."

Either is that.

", I think if you want to run a windows server, you'll need the extra admins to take care of the additional crashes and glitches, "

Is there some sort of three strikes rule?

Re:This isant what you think

[sheldon]

Well actually the IDC figures are also supposed to be 24% of server shipped.

The 8.6% sounds more likely.

Yes!

[Tofu]

I agree completly! So many IT problems are solved making decisions based on what is "used" or what is "popular" and not on what will solve the problem. Haphazrad solutions like this increase the problems complexity time and time again.

Re:Yes!

[Tofu]

Sorry,

s/Haphazrad/Haphazard

And here is the crux of the matter

[Loundry]

Some would argue that for such a large scale installation we should have vendor installed servers. But we found that we wanted a very customised solution, and without putting the vendor down, we felt that we had enough expertise within. At the end of the day we were in complete control.

And therein lies one of the undeniable powers of free (libre) software: the user is in complete control.

This is something that proprietary software cannot compete with. Proprietary software always takes some control away from the user. With free software, the user always has control.
Advocates of proprietary software may not like this fact, but it doesn't make it a troll.

Re:Pre-installations were not involved in the numb

[Rasputin]

1. Deliberately hurts Mp3
unless you use 3rd party software

But that's exactly the point. In the Windows world, requiring you to go to a 3rd party solution is the kiss of death.

Join me in the Microserf chant: "If it wasn't sold by Microsoft you can't trust it to work with Windows".

Gartner says water is wet (0.8 probability)

[gelfling]

Oh pulllllleeeeeeze !!!!!

Gartner has a long and glorious history of being a shill for MS and for the patently stupid. For chrissakes they invest in the companies they "study" and the managing directors also sit on the board of a VC firm that pours money into IPO companies that have "new and amazing technology on the brink of greatness !!!!"

And even if they weren't unethical criminals how often have the actually been right? Um....I'd say about a 0.15 probablity so to say or about 1 in 7.

Re:Lets talk about "recycled" servers

[astrashe]

I think that this is an important point. The idea of studying what happens when someone "buys a server" tends to push you towards MS.

I'm sure we beat MS in recycled servers by a wide margin, though. :)

Re:Pre-installations were not involved in the numb

[elflord]

I simply do not believe that Linux is shipping on 25 percent or more of all new servers and I just cannot believe the IDC figures

This is not at all inconsistent with the assertion that he actually asked customers what they were running -- the point is that you'd expect that the number of preloads would exceed the number of users, or at least be not too much less than the number of users.

If the majority of customers got their Linux via download or some other way, "then the market is in even worse shape than my survey shows,"

Right -- because it means that downloads are getting counted in his survey, but they don't show up in sales figures, and download users are also unlikely to purchase support contracts.

What is the definition from a Windows server?

[Julz]

Hmmmm. Do they/you class a windows machine infected with say the IRC bot that attacked grc.com recently as a kind of server, or any of the other BO trojans/servers, Netbus, etc. How about all the windows boxes out there sold as a home system with file sharing setup under TCP/IP and no firewall like ZoneAlarm protecting them. Surely that could also be classed as a type of server, for better or for worse. We know which one!
Heh if you looked at all those and all the misconfigured machines out there, then there probably is a higher number of Windows servers out in the wild. Especially with the high number of flatrate ADSL, ISDN and Satellite services that only offer Windows drivers on the setup disks.
What can you do?

Re:NT "Cost Efficiency"

[ergo98]

This is not to say that NT goes down every five minutes. But in order to keep one NT server up even approximately as long as a Unix machine, I find that I must restrict it to doing just one service at a time (i.e. just mail, or just file serving, or just web, or just DB, etc.); whereas on a Unix machine I can frequently roll several services onto one machine with no significant drop in performance or reliability.

Nonsense. While there are some applications that tend to like the PC all to themself (MS Exchange being once such product, or SMS being another), you can always put SMTP, FTP, WWW & SMB file sharing on an NT / 2000 box with zero effect on stability (presuming you keep up on service packs). DHCP and DNS are also no brainers on W2K machines. This whole bullshit about Windows only being able to do X number of things versus Linux doing X^16 is due to the fact that in the land of Linux things like file sharing are a big deal and it's a big deal if you have Samba running, or WU-FTP, etc, whereas having the server running in NT/2000 is a no-brainer, and most people install the inetinfo services without a second thought. Same goes for DNS, DHCP, etc. etc. etc.

The only issue I have with Windows machines is that they are far too used to putting the "make the machine reboot" flag in their install files. It pisses me off when I install some trivial file that should at most require me to restart a service but instead I have to restart the PC.

Re:Think about this...

[ergo98]

Lets face it, my 16 year old sister can learn to be a windows admin from one of those tech schools that'll make you certified in 3 months. And you can pay 5 of those newbie admins for the price of one unix admin.

Ah the delusion of UNIX elitism. Yes UNIX is hard and Windows is easy. Oh wait: There are millions of 14 year old boys and girls who've downloaded Linux and spent a couple of months plugging away who now consider themselves experts (hence Slashdot)...gosh, almost seems like your hypothesis is ridiculous.

I would wager a pretty good penny that there are far more "experts" of UNIX out there than there are "experts" of Windows. It isn't cool to know how to delegate authority in Active Directory, or to set up secondary zone transfers for multiple tree domain forests, but it sure is cool to know cron.

Re:NT "Cost Efficiency"

[ergo98]

The PDC is also a webserver, RAS server, proxy server, FTP Server, and SMTP relay. A BDC is running web services, numerous internal services, FTP, telnet, netmeeting RDS, SMTP, SQL Server 7, file sharing, and of course the standards such as IPSec. On both of these reboots are measured in months between and have been the result of new SPs or critical hotfixes.

With Windows 2000 you can easily set up a domain Active Directory domain controller, DHCP, DNS (AD integrated), FTP, HTTP, SMTP, LDAP (AD integrated), SMB (file sharing), IPSec, Certificate Authority, SQL Server 2000, presuming you have it on proper hardware (for example a good amount of memory, and given that memory is about $100 for 512MB that seems like it isn't asking for much).

The illusion about only being able to run X amount of things on NT/2000 is the result of people reading MCSE books that are geared towards very large corporations (i.e. stand along DCs) and applying it to tiny office corporations...it's simply misinterpretations of what they're saying.

Re:NT "Cost Efficiency"

[ergo98]

I guess you don't bother to follow Microsoft's list of "Best Practices", then.

Best practices and real world practices are two very different things. When running an intranet site the security concern is negligable, and the same security practices apply to running ANY software on any system with anything remotely confidental. If everyone followed that WU-FTP, BIND, Apache, etc., would all run on separate servers, so this whole argument would be moot.

As for having Windows 2000 share that much, you probably don't want to set it up that way. The way MS SQL Server was designed, it will take every last byte of memory and can be a real hog for CPU cycles if it has work it believes it has to do.

By default SQL Server takes all available memory (slowly gaining as it doesn't release cache memory), however it returns it to the heap when other processes start requesting it, and secondly there is a slider (as well as of course system settings) that lets you set the starting and maximum memory: Problem solved. SQL Server is a user mode process and it gets standard task scheduling just like every other process, so barring setting it to high priority (there is a setting for that if one does set up a independent machine) it doesn't take over the machine. Does it want all it can get? Sure it does. Personally I have it set to a one-processor affinity on a two-processor machine so there could never be a problem anyways.

Re:NT "Cost Efficiency"

[ergo98]

In a previous post, you mentioned that being able to stick all these services on 1 or 2 boxes because the load on them wasn't very high. At the risk of sounding like a zealot, with the money saved on licensing, an Open Source solution would allow you to invest that money in another two or three boxes, which would allow you to spread out the services like you should.

The whole point of this entire thread was the assertion that the TCO of Linux is less because you can do so much on one box, versus only being able to do "1 thing" on each Windows box. If people want to rant about the security IIS (BTW: Which is ridiculous. While I'm sure there's some black cloaked organization that was exploiting this before it was popular, 99.999999999% of those exploited were lazy fools that didn't keep up on system patches. Patch comes out->System is patched. The idea that internal employees are going to be haxxoring is laughable. BTW: The unicode exploit ran under IUSR_Machinename -> An account which on properly configured machines has ZERO permissions. Unfortunately though too many people are aware of things like security templates [because they're lazy]), then my point was that the exact same consideration should stop you from putting any 2 services on a Linux box.

Whatever OS/apps you choose, the services *must* be segregated to separate boxes to reduce the risk of harmful 'interactions' and to spread the risk associated with hardware failure. An admin who fails to do so is just being negligent. If an admin is being forced to do so by mgmt, then the admin has failed to properly present the risk/rewards of current scenario to management.

How simplistic. Say you have a DNS server, web server, and SQL Server, and you have the choice between 3 regular PCs, or one RAID 5/ECC memory/dual-power supply/dual-processor (SMP but with processor failover in case of an issue), dual NICs in a redundant role, etc. Which is more reliable? In my experience the second is of course many magnitudes more reliable, but it also reduces TCO by requiring 3x less maintenance for security patches, etc.

Re:NT "Cost Efficiency"

[ergo98]

I'm not saying it's harder on NT, but man, I wanna know what you've been smoking to think it's a big deal on linux.

Sorry I realize I was rather vague, but my point was not that the installation process is harder, but rather that the perception of what the machine is doing tends to be different in Linux land. i.e. My machine is running WU-FTP and BIND! In the land of Linux individual services seem to have more mindshare, whereas on an NT/2000 box they're often just a given. No one makes a big deal about having the Server service running (hence SMB file sharing) on 2000, but it's a huge deal if it's SAMBA (again perceptually). My point was merely that people load down an NT/2000 server (I wish there was a catch all name for them) with services, but because it was all from the box it's a "clean" server (despite having a huge gamut of services) that is "not running anything".

When are these guys going to learn?

[rnturn]

Gartner, IDG, et. al.: It doesn't matter how many are shipped with Linux pre-installed. To count market share that way omits all the copies of Red Hat, Suse, etc., that are purchased directly from the distribution manufacturers and then installed on gawd-only-knows how many computers. I know... I purchased one copy of a distribution and installed it on at least ten systems. I don't know on how many computers that copy of the distribution was installed by the people that borrowed the CDs from me. Let's not even get into the number of systems out there running Linux that was installed from a CD burned from a downloaded ISO image.

Get off the idea that market share is tied to pre-installations. Please. You just look stupid in the eyes of the people who will, someday, be running IT shops. Those people will remember your faulty methodology and look elsewhere when they want to do industry research. Your expensive reports will merely gather dust on the shelf.
--

Re:NT "Cost Efficiency"

[Fyndo]

that in the land of Linux things like file sharing are a big deal and it's a big deal if you have Samba running, or WU-FTP, etc, whereas having the server running in NT/2000 is a no-brainer, and most people install the inetinfo services without a second thought. Same goes for DNS, DHCP, etc. etc. etc.

Huh? "apt-get install apache".

what a big deal that was. Really. I'm terrified of doing "apt-get install samba" now.

Christ, Redhat installs too damn many daemons on every machine. FTP a big deal? I have to go to new installs and turn it off.

I'm not saying it's harder on NT, but man, I wanna know what you've been smoking to think it's a big deal on linux.

Re:Pre-installations were not involved in the numb

[jelle]

"I wonder if they pick randomly from the companies which buy their reports."

I wouldn't be surprised if they used the customer lists from the companies.

Re:Probably about right for MARKET share

[kinkie]

At Linux Expo Milan last week there was the Linux-Mandrake CEO delivering a keynote speech, and in that he stated that Linux Mandrake is very close to profitability.
A few weeks ago it was reported here that Red Hat is very close too.
If that is correct, I don't think either is going belly-up any time soon.

Re:Around here...

[sharkey]

Well, la-dee-da! I'm running NO ROM BASIC.

So there!

--

Re:24% Ho ho ho!!!

[Felinoid]

Chances are fairly good Linux shops don't need/want your services.
Keep in mind management knows Linux isn't a real Unix. Calling you guys might seem as silly as calling a Windows expert to for a Mac.

I agree on one point. A server is a fuction. But it is likely they are mesuring speed not function. That produces a poisoned result.

I would say your client list is also a bad sample.
There isn't a market for Linux consultents like there is for Windows and Solarus consultents.

Linux consultents are more like Mac consultents. Rarely needed.

Also the sereousness of the need will tend to dictate BSD or Solarus over anything else.
But thats not bulk but importence...
For importent stuff BSD is the best choice.

Study automaticly bies

[Felinoid]

Forget who paid for the study. It could be Andover.
The problem is what the study is.

In effect they are counting the amount of preinstalls. Linux is mostly installed after market. Windows is mostly preinstalled. So Windows is automaticly going to show up the better.

Also Windows requirements increase over time forcing new machine sales just to use the latest updates. Linux requirements don't incrase so Linux sales only relfect the need for better hardware.

At most maybe Gartner set a higher standard for "server" than IDG did.. Linux usefulness starts to fall off at a certen point on the high end so sales of extreamly high end is likely to be Windows NT only.

The reality is however that for Linux a 386 is a server but to Windows nothing less than a Pentium Pro.
But you can only mesure "server" by CPU power. But what makes it a server is the intended use. That is an unknown.
A bunch of Win 2K server boxes sold could be for workstations or game machines. A bunch of the Linux servers in the sample could be destend for a Beowolf cluster.

It's a self poisoning servey...
I would presume to say that Linux is under represented. But being truthful... Linux could also be over represented.
It is a random number of no meaning.

Re:Ok...

[Midnight Thunder]

To a good techie this definetly has a high "duh" factor. Now take your average management type (see Dilbert for explanation) who reads these articles and you will see them taking this very seriously. When the management type talks to his/her techies you can imagine the prevailing argument as the enlightened explain this to the decision maker - you'd be surprised how many unqualified people make decisions in place of the qualifed.

Wait a minute -

[wirefarm]

Take a typical shop with let's say 10 servers.
If they run Red Hat on all 10, they probably only have one copy of Red Hat on CD, possibly borrowed from a friend.

OTOH:
If they run NT Server on all ten, they probably have... Oh, wait... Never mind. ;-)

Cheers,
Jim in Tokyo

MMDC.NET

Depends..

[nyet]

Is this report targetted at os/server vendors or os/server purchasers?

Seeing as the report indicates that PURCHASES of linux are down, that shouldn't affect your decision to BUY, given that you understand that the raw number of purchases in no way indicates a) popularity and b) (by a fairly flawed populist metric) stability/suitability.

However, if you are a VENDOR, then all you care about is net profit anyway, and you don't care whether Linux or Windows is *installed*. You ONLY care about how much money is *spent* on them. If you are a VENDOR you LIKE large TCO as long as the resources spent on that big TCO are going to you.

Microsoft commisioned this report as a VENDOR, but released it with the hopes that it will discourage VENDORS as well as BUYERS to avoid Linux.

Everybody is so busy whining about the Gartner vs IDC numbers that they are failing to see the forest for the trees.

Re:Slashdot should not assume its readers are MORO

[Black Parrot]

> For one who goes to lengths to bash corporations and greedy managers, this statement sure doesn't make you look like one with much of a backbone. Imagine--a whole career spent at places where you didn't respoect the management. Maybe the safety of a paycheck kept you from leaving a place that was philosophically dissatisfying to you?

Oo ow ouch, that stings!

Oh, wait a minute. I haven't had any managers looking over my shoulder for almost eight years now.

--

Re:Slashdot should not assume its readers are MORO

[Black Parrot]

> Gartner is very well respected in the industry.

Respected maybe, but clueless all the same. Gartner has a long track record of "predicting" for next year exactly the stats that others found for last year. I haven't seen their secret business methods, but I rather suspect that they just collect whatever stats they can get (which will in general reflect what was happening a year ago), change "stats" to "predictions", and sell it to gullible managers... like you.

> To claim this is biased because Microsoft funded it is absolutely ridiculous.

Or not. MS habitually funds studies with a little string attached, which says "You cannot publish the results of this study without our approval."

That doesn't result in an accurate view of what's going on in the world. In fact, it's essentially identical to the "science" practiced by dishonest or self-deluded paranormal researchers, who publish their result from trials when they get lucky, and ignore the results of unlucky trials on the grounds that "the force wasn't with me today".

This kind of thing makes it really easy for people to justify believing whatever they want to believe... whether they be paranormal researchers or corporate managers.

> When my corporation pays for research, we absolutely do not want them to tell us what we want to hear.

I've worked for a number of corporations, and never once met a manager who wouldn't tune out what he didn't want to hear. Most would only pay for information that furthers their careers; anyone who staked his career on converting his company's Sun boxes to Windows boxes will be more than happy to pay for this particular report.

> W2K IS a serious Enterprise-ready scaleable and reliable OS.

Gosh, I'm down to posting a link to the Hot 100 analysis almost once a week anymore. I'll let you find it yourself this time: go to groups.google.com and search comp.os.linux.advocacy for "hot 100 uptimes", and read what you find. Or do a new analysis yourself, and post your results for us.

> Just because Microsoft commissioned the research does not render it invalid.

Repeating it over and over might make you feel better about the money you wasted on the report, but repeating it will not make it come true.


--

MS isn't all evil, but you aren't making a single

[sweetooth]

point in their defense. I dunno about the rest of the people out here, but the whole "ease of use" of windows is BS. Windows is easy to use once you've used it over and over again. Take three people with no computer experience. Sit one down in front of a Mac, one in front of MS Windows 2k, and the other in front of a linux box with KDE 2 or Gnome or even E. Now, train each of those individuals on how to use the tools in front of them for the next 8 hours. At the end of this you'll find that they are going to have an easier time using what they have the 8 hours of training in than the other OS's. Certain aspects may carry over, but for the most part "ease of use" is nothing more than familiarity. I know people that hate GUI's cause they've been using command line tools for so long, they find the gui's to be harder to use than the command line. Once again, it's all familiarity and not "ease of use".

Now, not all of MS is evil, a portion of thier business practices are very questionable, and when you are as big as MS is you get generalizations. Look at how many people hated IBM (Big Blue) when they were on top. If you want to break down and find out what's good about MS, go take a look at thier hardware, I'm sititng in front of an original MS Natural keyboard (with the full size keys and the lift in the front, none of that Elite crap) and a MS Intellimouse Pro. It's excellent hardware. Look at the games the MS has published (note published and not developed) such as Age of Empires and Asherons Call. Whoever is running publishing appears to know what they are doing.

I think I'm just arguing that yes, MS isn't all evil. However your post doesn't point to those things. Ease of use isn't one of them, and even if it was they stole it from Apple which can be considered evil.

Re:Is having MORE servers with your OS really bett

[cybrthng]

Since when did/does it take more NT servers to do what linux does?

Shoot, my pc at home runs Windows 2000 server and Oracle 9i with a hacked up 9i Portal as well. I have DNS, Sendmail, Telnet running and still have cpu to play unreal tournament, read my emails and do whatever else i want to do.

Sure linux can do that, but i'm using the same hardware linux would run on and well, NT/2k does it just fine.

I setup tons of application servers being a consultant for a big firm. I have never once seen a shelf of NT servers doing what "one" linux server can do..

Infact with all the beowolf hype, people in the business land think you need 4 servers to do what 1 nt server can do. People know NT/2k can run Exchange, people know NT can run DB2, Oracle 8/9i and Oracle Applications 11i.

Linux is still proving what it can do.

And doing a damn good job at it.. but alas you don't need more nt boxes to beat a linux box and whoever spits that out is just a moron.

If it's not an independent study...

[Mdog]

...why do you people even pay attention to it? I'd say that giving it as much attention as its getting is a win for MS.

duh

[Mdog]

That would be bad too. It's especially worse that MS has a reputation for unethical business practices. If you don't find that reputation to be true...go back to zdnet :)

Re:Pre-installations were not involved in the numb

[VP]

Also according to Gartner:
Some 724 U.S.-based respondents had answered questions over the phone for the survey, he said, ranging from small organizations with fewer than 50 PCs to large companies with more than 500 computers, as well as educational institutions, Internet service providers and application service providers.


There is no indication of how these respondents were selected, and how representative they were. I am sure anyone can pick 700 respondents who will show 100% Linux or BSD loaded on new servers.

Re:Gartner smells like Ziff Davis

[bgarcia]

As has already been stated, Gartner asked end-users what they installed on their computers after they bought them. Not what was pre-installed on their computer. Implication: either Linux doesn't have the marketshare zealots want to belive, or accurately assessing server marketshare is difficult. You decide.

But then, here's a quote from Gartner's survey:

The study results indicated that in the traditional server market in the United States during the third quarter of 2000, 8.6 percent of server shipments were Linux-based systems.

So you might want to try reading the survey itself, instead of some marketing guy's <cough> misunderstandings about it.

Re:Hmmmm

[thrig]

Checkout netcraft, they seem to be doing a good job of tracking that sort of thing:

http://uptime.netcraft.com/up/graph/

Re:We need a FUD log

[Malcontent]

Zdnet, devx, microsoft.com asphole etc are all very heavily pro MS. For every slashdot there are a hundred MSdots.

Re:Think about this...

[Malcontent]

I think you nailed it. Knowing windows is not cool. Linux is cool. Fortunately for us MS actually hypes this aspect of NT. They are always talking about how you can get by with cheap admins on NT vs paying for expensive Unix Admins. I keep a list of these clippings and whenever a high schooler asks me what he or she should learn I tell them that Bill Gates or Steve Balmer says they will make more money by being a Unix Admin. Ain't irony sweet.

I would suggest this to everybody. Go dig around the net and gather all comments from MS executives about how much Java programmers cost or how much Unix Admins cost and then go visit a high school. You can make a huge difference in someones life.

Re:NT "Cost Efficiency"

[Malcontent]

What you'd share your PDC with other apps? Your SQL server? Most NT shops I know keep separate machines to do nothing but PDC and SDC.

Probably about right for MARKET share

[throx]

What Gartner/Dataquest is probably trying to measure here is the marketshare Linux has - in other words, how much money is flowing through each section of the operating system industry.

Thanks to the wonder of open source software, it doesn't really matter if 75% of the actual machines are running Linux (and most machines running Linux are probably assumed to be servers), if they are all downloads with no support contract then Red Hat, Mandrake and the others are still going to eventually run out of cash and go belly up.

What these numbers from IDC and Gartner are essentially saying is that Linux itself may be Red Hat's biggest competitor. If RH is only getting 10% to 15% of the money coming back to it then they are most likely being cut off by the very free (beer) nature of the GPL that allows you to install as many times as you like.

The other interesting thing from the ZDNet article is that no numbers for NT/2000 were mentioned from the Gartner study. I wonder what the split really was among the other operating systems?

Re:Probably about right for MARKET share

[NutscrapeSucks]

The other interesting thing from the ZDNet article is that no numbers for NT/2000 were mentioned from the Gartner study. I wonder what the split really was among the other operating systems?

The general way this works is that the headline and the executive summary are free, the details cost $$$.

Re:This isant what you think

[mfarver]

Actually.. I noticed the apples/oranges comparison and figured most of /.'s readers would too (maybe I was an optimist). What's interesting is that both the article and Gartner seem to be making the argument that their findings prove the IDG study wrong.

I can tell you our office doesn't purchase servers to run linux. Instead linux is running on all the early pentium machines that Windows 2k won't, and quite well at that.

Re:Gartner smells like Ziff Davis

[cafeman]

Three points.

1. As has already been stated, Gartner asked end-users what they installed on their computers after they bought them. Not what was pre-installed on their computer. Implication: either Linux doesn't have the marketshare zealots want to belive, or accurately assessing server marketshare is difficult. You decide.

2. Garter is a research agency that has its value locked in its reputation. If it produces poor quality research on a regular basis, people will stop subscribing or purchasing its reports. Short term gains (ie fabricating truths in reports for specific companies) will lead to long term losses (ie loss of reputation leading to loss of credibility and revenue). Are you arrogant enough to believe they haven't thought of this? The reality is, they analysed the information available to them and made the best, least biased forecast as they could.

3. Garter is not in the business of assessing obscure technical facts. They provide a strategic business perspective on technology. Tech-heads are not their market. They don't care about the operational aspects of the technology. The people who run business (which, most commonly, are not tech-heads and have different skills) are their market. They care about the strategic implications of the technology and longer-term market trends.

   You think you could do a better job and produce higher quality research? Don't whinge - go out and do it. If you produce research of such a high calibre, you'll drive Gartner out of business in months. I'm constantly amazed at how many people on Slashdot are willing to spout off on things they know nothing about. I don't tell people how to fly a plane - I realise I don't know how. But, the village idiots on slashdot are always willing to provide legal advice, assume everyone in management is a PHB, that companies never ever know what they're doing, that everything is part of a conspiracy and that anyone who doesn't know how to write a sound compression script using bash is an idiot. On the other hand, some people here have some seriously big clues. You my friend, are not one.

In summary, read the bloody report and get some perspective before spouting off.

easy answer

[nuintari]

$lynx -head -dump http://www.censorware.org/ | grep Server
Server: Microsoft-IIS/4.0
$echo your welcome

means NT as the OS, since win2k runs IIS 5.0 afaik.

absurd

[Illserve]

What a perfect example of a steamy moist pile of bullshit. As rob points out, an overwhelming number of linux boxes are probably installed overtop of the grave of a windows installation. You can't buy a computer *without* windows from many of these places.

Then there's the homebuilt faction, and I'm sure a hugely disproportionate number of linux servers as opposed to windows boxes are made from mobo's and CPU's.

This is a textbook example of how not to do statistics.

Thanks MS.

Re:absurd

[TrollMaster3000]

Actually I find that MOST Unix/Linux users build their own machines. I only have a PIII that I bought from Compaq, with a blank HD, and put Debian on it. Runs fine, been up over 230 days.

Re:absurd

[xandria]

Hey dummy -- read the Gartner piece. It specifically said that the respondents were asked what OS was (or will be) running on the purchased servers, not what OS was shipped with their server. Gartner's Hewitt said:

"We went to end users, rather than looking at just sales numbers, and asked them what servers they had bought over the past three months and what operating system they had installed on it over the same period," Hewitt said. "There was no question about whether Linux was preinstalled or not, we simply asked about new shipments and this is what we found."

Re:My server Runs linux

[questionlp]

When I got my server at home, I needed something that didn't occupy a lot of space and really didn't need it to be beefy at all. I decided to get a Compaq iPaq desktop, but Compaq only provided Windows 95/98 or Windows NT with it. I got the one with Windows 95, wiped it, and put FreeBSD on it. You can read the little paper on it here.

Re:The big boys sell machines with linux... poorly

[mozkill]

dude, that is a awesome story. i can totally see that happening. isn't it crazy? it makes me wonder about the guys that work at Dell whom make the hardware decisions... they must be smoking crack. actually, my guess is that the choice of hardware in those "Redhat" machines was made by a salesperson, who got a good "deal" for Dell on the hardware. Sorta like how Compaq and HP work. :-) You gotta know that the engineers there probably know what there doing, but its the salespeople that choose the hardware, and that is almost pure luck of the draw.

why we all post to Slashdot

[mozkill]

I love all the "unix love" that goes around on Slashdot. It makes me realize why we are here, and why I love this site so much. I just wish the Microsoft people would get out of town, just like I wish the Californians would stop moving to Portland. :-)

Re:We MUST quash this high tech economy!

[be-fan]

Yo, stop mistating communism. (I don't care if you're being sarcastic, either way you're misrepresenting communism.) Have you even read the Manifesto? I don't agree with some parts of it, but in reality, much of what the US is now is basically what Marx outlined in the Manifesto. Too many Americans get their ideas about communism from cold-war type idiots, and dislike the entire ideology based on Russia's basterdized version of Communism.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Comment removed

[account_deleted]

Comment removed based on user account deletion

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:Pre-installations were not involved in the numb

[Ded Bob]

There is no indication of how these respondents were selected, and how representative they were. I am sure anyone can pick 700 respondents who will show 100% Linux or BSD loaded on new servers.

I wonder if they pick randomly from the companies which buy their reports.

Actually, the article probably avoided the answer because it might have calmed down the flames and therefore the number of hits on the page. Advertising rules all.

Pre-installations were not involved in the numbers

[Ded Bob]

At least according to Gartner:

Respondents were screened to ensure they were knowledgeable about server purchases over the quarter, and they were asked what percentage of their server purchases were Linux servers, he said.

"We went to end users, rather than looking at just sales numbers, and asked them what servers they had bought over the past three months and what operating system they had installed on it over the same period," Hewitt said. "There was no question about whether Linux was preinstalled or not, we simply asked about new shipments and this is what we found."

Just Microsoft is involved does not necessarily mean the numbers are bad. It is hard to say.

Re:RTFA (Read the Friendly Article)

[blazerw11]

Gartner said that they asked end users what OS they had installed on their new servers

Even you are claiming that the survey is talking about preloaded OSes. Who installed the OS if they "had" an OS installed?

Did they count the mainframes?

[jlrowe]

Mainframe computers are now running Linux, and did not come pre-installed that way. More significant is that one mainframe can run thousands of copies of Linux.

IBM has changed from whatever they were using before to Linux on their home page. Hard to say if it is running on Linux, but it sure isn't running Windows.

lynx -head -dump http://www.ibm.com | grep Server
Server: IBM_HTTP_Server/1.3.12.1 Apache/1.3.12 (Unix)

They didn't count me....

[SwedishChef]

and my company specializes in replacing windows servers with Linux servers. In the past six months we have:
1. Maintained one school district running all Linux... costs them $400 a month (3 servers);
2. Installed a RAID Linux file server in a small lawyer's office... 2x30GB turnkey for $1200 hardware included;
3. Replaced a winNT3.51 file server for another attorney, converting all his files... $700 using his box/cpu;
4. Installed DNS/DHCP/Mail server running Linux for a local ISP using an old dual-PPro board they had laying around... cost $250;
5. Three NAT servers for small businesses using Squid proxies... converted from old 486 machines for $200 each;
6. 4 Lineo NAT boxes (the Moreton Bay units) for clients running fiber from the local PUD... cost about $400 each to the clients installed (running VPNs);
7. NAT/Squid for local governmental agency which finally got ADSL to work... another conversion for about $300;

That's seven installations from one SuSE 7.1 box plus the Lineos through May from one, 2-man shop in a small town in the middle of nowhere. Everyone counts the big boys... but it's the grass-roots where the revolution is taking place.

Re:Gartner smells like Ziff Davis

[haggar]

I am not sure IDC smells any better.

Being so involved with high and midrange servers, I was very suprised that IDC claimed 24% of servershare belongs to Linux. With all due respect, and indeed acknovledging that Linux does have it's functions in corporate networks, I believe that Gartner's figure is much closer to reality.

My 0.02, anyway.

Than

[Kidbro]

Gartner Claims Less Linux Than IDG.
Please....

--

Re:So?

[KenSeymour]

It matters when they stop writing software for the OS you are using and you failed to notice that
everyone else stopped using it 3 years ago.

If you are an IT shop. Suddenly, you can't get any programmers.
All you good programmers are quitting because they are adding
years of experience using stuff no one else is using.

Maybe the software you are using on OS/2 is the best today.
But what if no new versions come out for it.
Meanwhile the Windows solution is heavily developed
because they are getting the sales and can plow the money back
into improving the software.

It does matter how many other people are using the same OS.

Re:Gartner smells like Ziff Davis

[Shelled]

A few logical inconsistencies here. Point 1 implies that trends equate to results. This isn't necessarily true. One hundred times more Win than Lin servers could have shipped in the last quarter and Linux could still have a 25% share. What matters is long term trending.

Point 2 is just wrong. I work in an industry in which a not uncommon quip is "those who can do, those who can't become consultants." People who prove themselves incompetent after repeated failures in the workforce incorporate and sell their services to newcomers in the industry. A more familiar example are film 'critics' who make a very good living championing completely forgettable releases as feel good hits of the summer. People and companies willing to whore themselves are not in short supply.

Ah well, it's more fud.

[jidar]

More of the same. Anyone with any knowledge on the matter knows that Linux has far more installs than sales figures point out by a wide margin.

Netcraft is probably the best data we have on this, but even that is lacking.
Just ignore any information pertaining to the number of Linux installs, it is all just guesstimations, and yes that includes information that makes Linux installs look good.

Re:Ah well, it's more fud.

[xandria]

jidar, read the Gartner piece. It specifically said that the respondents were asked what OS was (or will be) running on the purchased servers, not what OS was shipped with their server. Gartner's Hewitt said:

"We went to end users, rather than looking at just sales numbers, and asked them what servers they had bought over the past three months and what operating system they had installed on it over the same period," Hewitt said. "There was no question about whether Linux was preinstalled or not, we simply asked about new shipments and this is what we found."

Ok...

[Moonshadow]

So...basically we're saying that the majority of computers shipped ship with Windows on them?

Wow, news flash!

The "duh" factor here is astounding.

Re:Ok...

[myster0n]

Wow, news flash indeed.
But on the other hand, the real news is that 8.6% of all new servers sold comes with Linux pre-installed.

This is in fact good news. Its only the spindoctoring in the Gartner study that makes it look either bad or not newsworthy.

Re:Ok...

[mikethegeek]

"Wow, news flash indeed.
But on the other hand, the real news is that 8.6% of all new servers sold comes with Linux pre-installed.

This is in fact good news. Its only the spindoctoring in the Gartner study that makes it look either bad or not newsworthy.
"

I think you make a great point there. Pretty much everyone who sells servers these days offers Linux. I work for one of the largest in Q/A testing, right now working on certifying their RAID cards and servers for Red Hat 7.1.

Linux is the ideal web server, and with SAMBA, an ideal file server for small and medium businesses. After all, connect ALL your users to it, no need to buy connect licenses... EVER!

Those categories represent, in raw numbers, the LARGEST bulk categories of the server market.

Re:Ok...

[xandria]

Hey dummy, read the Gartner piece. It specifically said that the respondents were asked what OS was (or will be) running on the purchased servers, not what OS was shipped with their server. Gartner's Hewitt said

"We went to end users, rather than looking at just sales numbers, and asked them what servers they had bought over the past three months and what operating system they had installed on it over the same period," Hewitt said. "There was no question about whether Linux was preinstalled or not, we simply asked about new shipments and this is what we found."

All this tells me...

[krmt]

is that statistics are unreliable at best. We've got conflicting numbers and interests all over the place, with all these surveys. In the end, it doesn't matter becuse Linux is a good server OS and it will continue to gain ground, especially as the younger people come up and start pushing it as an alternative to Windows et al.

What's really important isn't what numbers linux is doing now, but what it's going to be doing in the future.

"I may not have morals, but I have standards."

Who gives a wet shit?

[Legion303]

Even assuming Microsoft didn't pay for the study, and assuming the Gartner Group weren't a bunch of filthy whores, who cares if linux can compete with MS in the marketplace?

If you're worried that your PHB will force you to use MS products based on this study, well, find a job with more enlightened management, or take his job yourself.

-Legion

Linux Counter?

[gordzilla]

Have you registered your linux boxes at Linux Counter

Well, more now that Dell gives you a deal.

[sideshow]

They're selling a 800 PIII server with all kinds of goodies and software for 800 bucks! Wait! Having that software be Microsoft software costs another 800 bucks.

Re:Newton's law of slashdot posting

[zpengo]

I've noticed a trend with regard to this process. No matter how passionately the stoical marketeers defend Microsoft, nor how predictably Linux' yes-men define the particular news story as the turning point in the eternal battle between the forces of freedom and the forces of evil, the truth will lie somewhere in the middle. What's depressing is that those who take the middle ground are morbidly few. Who would have thought the tech sector would create such starry-eyed romantics (as many online activists seem to be)?

That's because only extreme viewpoints gain karma. I usually post on both sides of the fence for anything MS/Linux, and then watch them *both* get modded up.

that's liGNUx

[jon_c]

RMS actually wanted Linux+GNU to be the clever: liGNUx. however some people didn't think it had quite the "ring" it should have. so he decided GNU/Linux was good enough.

-Jon

Re:We need a FUD log

[Caspuh]

We should compile all this stuff and put it in a central location so that we can refer to it at a later time.

So you want a Microsoft version of Slashdot?

Re:Pre-installations were not involved in the numb

[jamesl]

This study was done primarily to provide information to industry participants, not the rest of us. The numbers developed can be used to project future revenues to companies that provide support for various flavors of servers. This is why it was designed to learn what is installed and running on machines, not what was on them when they shipped.

While it is nice to get paid for software each time it is shipped on a machine, the real bucks come from support and service contracts. It may also be useful for people making important career decisions about what software they are going to become expert in.

Re:Newton's law of slashdot posting

[jamesl]

And exactly 12% of the posters will have actually read the article referenced.

Re:NT "Cost Efficiency"

[Trepalium]

I guess you don't bother to follow Microsoft's list of "Best Practices", then. Because of the fact you need to reduce the security level to run IIS on a machine that's either a BDC or PDC, your webserver can become quite a risk to the security of the rest of your network. MS recommends that webservers be run on a standalone server so that it can take advantage of the "local" accounts on that machine, rather than relying on a domain user.

As for having Windows 2000 share that much, you probably don't want to set it up that way. The way MS SQL Server was designed, it will take every last byte of memory and can be a real hog for CPU cycles if it has work it believes it has to do.

Re:We need a FUD log

[Mr. Sketch]

I bet it wouldn't take long for all the inconsistencies in their arguments to fall apart

Do they have any cohesion now?

8.6% but what does it mean?

[SeniorDingDong]

Some scenarios I can think of: 1) 8.6% is actually correct and our perspective is biased given that we have greater brain capacity. This site for instance is 100% linux, the company I work for has about 50%, and the guys upstairs maybe 10%, but the guys upstairs are IT-idiots, they really, really are. I mean, the term idiot is by no means inaccurate here; their IT-IQ 60; they're friggin dumbasses, hoo-boy! You can bet they don't read slash-dot, and there are probably (conjecture) a whole bunch more like 'em. 2) Linux based OS does not require the most powerful computer. As mentioned above, older computers are easily fast enough for the tasks for which linux is a desireable OS. So a strategy is buy a new computer, put Win on it, convert old computer that used to run Win to linux. This is mentioned in the linked-to rebuttal, but is this good or bad? Doesn't sound too good for Redhat. How much of the 24% vs 8.6% disparity does this account for? What for instance is the % of new computers that were bought to replace an old computer that wasn't quite fast enough? Given that linux is more efficient, replacing a linux based host would tend not to happen as often as an MS based host. Maybe everyone purchased their linux-oriented host in Q2, didn't need to replace it in Q3. Could a linux host double up on tasks? If the CPU tends to be under utilized due to better OS effencicy, why not? 3) MS apps by nature require multiple hosts. Now this is big-time conjecture I am by no means a Win sys-admin. Is it typical that a windows host is dedicated to one task or can it easily do many? I only mention it because it sure seems to be the case at the company I work at. Printer server host, domain administration host, back-up host, source-code-repository host, etc. Or is it a trait of windows sys-admins? 4) Economics led to purchasing slowdown. Q3 seems kinda far away from Q1 2001, but maybe things were already starting to fall apart then. Redhat's stock dropped, companys grew hesitant and/or all those .com early adopters of linux were running out of money and/or companys made due with the hardware they could and only invested in new machines when absolutely necessary given 2 and 3, this would be a preponderance of MS-based hosts. With so much left up to conjecture one asks what is the merit of presenting a number of 8.6% to the world. I can only imagine it's PR aimed aginst Redhat. I think I need to go buy a copy. Up till now, I've been fine with the free downloads, but really, after installing it on every single computer I own (except the mac) I think I owe them something. -Jeff

What's a PHB to do?

[MotoMannequin]

It is obvious (to us) that most servers running Linux are not "shipped" with it installed.

The real danger here is that some corporate manager will see this headline snippet, and react with something like:

"I knew this was a passing fad. Those geeks in tech don't know what they're talking about," etc.

Despite the misleading nature of these numbers, too many statistics like this will eventually add up to too many setbacks, and less acceptance. It's amazing the amount of people out there that have not idea when it comes to computers, but they want to believe that they do.

MotoMannequin

Re:What's a PHB to do?

[xandria]

Hey dummy: Read the Gartner piece. It specifically said that the respondents were asked what OS was (or will be) running on the purchased servers, not what OS was shipped with their server. Gartner's Hewitt said:

"We went to end users, rather than looking at just sales numbers, and asked them what servers they had bought over the past three months and what operating system they had installed on it over the same period," Hewitt said. "There was no question about whether Linux was preinstalled or not, we simply asked about new shipments and this is what we found."

Re:Who buys systems?

[swordgeek]

I believe they said servers.

The shit you (and I) glue together from random spare parts aren't servers. The surplus PCs and sparc workstations lying around aren't servers.

If you're running mission critical apps/services on these machines, you're in trouble. Call me when you go bankrupt.

24% Ho ho ho!!!

[swordgeek]

Honestly, the idea that Linux runs 24% of the new servers being sold out there is absolutely silly! 8.6% isn't much better.

From where I sit, as a professional Unix consultant, I see roughly 1-2% of "servers" being linux boxes, and even that's stretching the truth a bit. What are they calling a server here--any machine that runs DHCP?

I suspect that their definition of server is just as valid as defining a mainframe as anything that can cranks out "x" MIPS. It just ain't right!

So?

[mrdisco99]

Why is this important? Does it really matter what OS everyone else is using? If these reports are affecting your IT decisions, then you have bigger problems.

It should be about what solution is best for the task at hand. Every installation is different. If everybody implements your solution on Windows, but it looks like it's really best served by OS/2, which do you install?

+++

One point.

[(void*)]

You do have a point about slashdot posters being arrogant pricks, in general.

But this is about the Gartner group. The whole point about this thread is the credibility of this so-called leader in marketting research. You cannot prove a point here by pleading to that credibility. That would be circular reasoning.

So keep your mind open, and listen to the slashdot crowd for once. They are the ones who have to live with the consequences of purchasing decisions made by management.

Linux isn't even competing economically

[pnadeau]

The thing that no one in the media seems to get is that Linux isn't even competing in the traditional economic sense.

For instance, one CD distribution of Mandrake at my office spawned 4 Linux systems. Economically that's one 'instance' (one sale) of Linux realistically it's 4.

It's telling that the only way to compete with Microsoft is to abandon the market entirely and give away your product for free.

--

4 open source boxes, 6 sunboxen, 3 consumer base.

[buridan]

1 indybox 3 penguin computing. all 4 are on the internet. comparably, i've put 6 sunboxes running solaris and 2 apple boxes and 1 winders box.

NT "Cost Efficiency"

[StandardDeviant]

I've administered both NT and Unix platforms (including the pricey proprietary ones). Unix wins on TCO, in my expereience, and it's no contest if it's x86 hardware unix.

Why? Business requires a certain minimum level of functionality from their servers, some level such that the temptation to go back to paper and pencils isn't a factor, which varies from business to business and from usage to usage. In order to reliably meet a given level of functionality (performance or stability), I have found that pretty much without fail NT required more machines and more powerful machines (== more expensive machines) to meet that limit than Unix did. More machines means more admins. More machines and more admins mean more cost.

This is not to say that NT goes down every five minutes. But in order to keep one NT server up even approximately as long as a Unix machine, I find that I must restrict it to doing just one service at a time (i.e. just mail, or just file serving, or just web, or just DB, etc.); whereas on a Unix machine I can frequently roll several services onto one machine with no significant drop in performance or reliability.

I am not an Open Source zealot. I am a pragmatist, and I only evaluate the tools I use based on how well they meet my or my client's needs. Microsoft simply does not provide good tools to run mission-critical services on, however "cost effective" they may be.

--
News for geeks in Austin: www.geekaustin.org

Re:NT "Cost Efficiency"

[skuenzli]

Best practices and real world practices are two very different things. When running an intranet site the security concern is negligable, and the same security practices apply to running ANY software on any system with anything remotely confidental.

You've got to be kidding me. What administrator would risk their PDC for the IIS vulnerability of the week/month? I think you're forgetting that something like 70% of cracks/hacks/security breaches are done by current/former employees.


If everyone followed that WU-FTP, BIND, Apache, etc., would all run on separate servers, so this whole argument would be moot.

In a previous post, you mentioned that being able to stick all these services on 1 or 2 boxes because the load on them wasn't very high. At the risk of sounding like a zealot, with the money saved on licensing, an Open Source solution would allow you to invest that money in another two or three boxes, which would allow you to spread out the services like you should.

If you think software/hardware is costly, then run the numbers on stopping the work of an entire company dead in its tracks because a disgruntled employee exploited an unpatched IIS Unicode vulnerability on your PDC/web/ftp/file server (let's go ahead and format the shared drive while we're at it--we've got a remote shell, after all; just to make sure we make the end-user impact measured in days/weeks of lost time)

Whatever OS/apps you choose, the services *must* be segregated to separate boxes to reduce the risk of harmful 'interactions' and to spread the risk associated with hardware failure. An admin who fails to do so is just being negligent. If an admin is being forced to do so by mgmt, then the admin has failed to properly present the risk/rewards of current scenario to management.

Stephen

Re:NT "Cost Efficiency"

[NutscrapeSucks]

Well, if you run IIS on the domain controller, you have the obvious problem of computer\IUSR_comptuer becoming domain\IUSR_domain, but that can be worked around.

In most networks, the Primary domain controlling is not a very resource-intensive application. The BDCs handle the business of communicating with the workstations, and the PDC only needs to replicate the SAM every 40-or-so minutes. So, it is the perfect box to load up services on.

IIS, on the other hand, has far too many reboot scenarios to use on a box is directly relied upon for user work. That means, just practically, no file servers or client-server servers. So, if you are firewalled, why not the PDC?

As a user application, SQL Server is well behaved as you say. Historically, the problematic stuff is the Microsoft kernel-mojo software, such IIS and the file server. And Exchange, because it has to spend lots of time fighting with it's crap JET datastores.

Re:NT "Cost Efficiency"

[maxpublic]

It's been my experience that MS servers require more maintenance than Linux/Unix boxes, which means more time and more admins. They also tend to choke more often when under load. Just my experience, mind you, but it's in contradiction to yours.

There is one thing that MS can never beat Linux at, one thing that really appealed to the bean counter in my employers: Linux costs nothing and has no licensing fees - ever. Even if Linux and MS products run head-to-head in terms of their viability as servers in real-environment conditions, the cost savings tend to stack up over time. Why pay for a product when the alternative is just as good and free?

And don't go into 'service contracts'. I've never received any remotely useful information from MS for fixing a problem, yet can get extremely knowledgeable assistance for free from Linux fans on the net. And I can rewrite the code to work more efficiently in a specific environment, something you *just can't do* with MS products (this has got to be the most frustrating thing for me - not being able to see the code and change it as I see fit).

Take those things into consideration and Linux doesn't have to be better than MS where server software is concerned - just as good will do.

Max

I bought one

[netinlet]

Dell Dimension 933 w/ 256 mb ram. Came pre-installed with RedHat linux.

It has never seen any other OS on it...

Doug

If Mexicans only make $500...

[RasTafarii]

how are they going to afford $24.95 for AOL?

or a pc running anything, let alone being able to read the man pages to install linux, since most of them are illiterate or left school at 3rd grade.

Here's another survey:

[big.ears]

This netcraft survey shows that out of the top 50 hosting locations (by uptime), representing 1754 sites, 24 total sites are running Windows derivitives (about 450 are running heavy-duty Unix offerings, and the rest are Linux/BSD.)

Not that this fact is particularly relevant, because perhaps the hosting locations that use Windows don't make the top 50 uptime slots. Seriously, though, what I'm pointing out is that there are a number of ways to skin this cat: IDG and Gartner have two different assessments, and I have a third. It wouldn't surprise me if we are all correct.

Re:Gartner smells like Ziff Davis

[__donald_ball__]

In summary, read the bloody report and get some perspective before spouting off.

Or you could try this experiment. Find some predictions Gartner issued a few years back and compare them with the state of the world today. I suspect their accuracy will be demonstrably poor.

I recall one prediction issued in the early part of 2000 - I think it was Gartner but could be wrong - that by the end of 2000, ~85% of Americans would be using wireless devices to access the internet. Anyone with half a brain should have known that was a hopelessly optimistic prediction - but these chumps manage to sell that kind of bullshit all day long. It's astonishing, really, when you think about it.

What the living fuck?

[bellings]

Jesus holy shit fuck christ. If a student submitted this study in any undergraduate research class, he'd be laughed right out of the damned building.

They use the world "methodology", and then have a paragraph that doesn't describe their methodology at all. They neglect to mention what their questions were, or how they selected respondents, or what the completion rate was, or what any of the results might have been. They say "Respondents were asked what percentage of their server purchases consisted of Linux servers." If that's what they asked, then how the hell did they get "percentage of redhat", or the total dollar value of the servers? If it's not what they asked, then why the hell won't they tell us what the did fucking ask?

The whole thing is a crock of steaming fetid shit, and anyone with an IQ over 95 would know it if they read it. The very best way to live up to the "lies, damned lies, and statistics" bullshit is to publish dick like this, where you refuse to say how the study was conducted (or even what, precisely, you were studying), and then cherry-pick a limited subset of the results. It stinks, and everyone at the Gartner Group knows it stinks, and they'll all go to hell for being such weaselicious fucks.

My server Runs linux

[BiggestPOS]

But I doubt it shipped that way, in fact, it didn't "ship" at all, we built it. These statistics are screwy simply because a LOT of linux servers are going to be home brewed/or maybe even shipped with Windows, and are NOW running Linux...

Oh, yeah... THAT Gartner Group...

[Skald]

I remember them...

• 1993: Windows NT (3.5) will have 80% of the desktop market within 2 years of release
• 1994: Internet will grow to 6 million users by 2005
• 1995: Cobol is used in over 65% of new application development
• 1996: Windows NT 4.0 will have 80% of the internet server market by 1998.
• 1996: Cobol is the world's 'premier language for application deployment' and 'there should be no worries about the viability of COBOL for any project on any platform'
• 1987: 75 percent of the Internet services for large enterprises will move to usage-based pricing by 2001
• 1998: '[Y2K] failures in less developed countries, smaller companies, and companies with high global dependencies will cause a negative impact to the world economy'
• 1999: Linux is 'Hype du Jour'. 'The lack of standards in the Linux community, coupled with a lack of key productivity applications and with Unix complexity, will continue to make Linux a poor choice for the mainstream business productivity user'. It soon appears that Microsoft sponsored the study. Gartner Group denies, but also quickly pulls the page from their site. Here's a biased synopsis

Re:Oh, yeah... THAT Gartner Group...

[shippo]

I worked for a Banyan reseller. When the sales started to dry up with the onset of NT, Banyan comissions a Gartner Group study that praised Banyan to the skies whilst knocking NT and Netware.

It was very PHB friendly, full of graphs and tables comparing various aspects of TCO - average number of users per admin, number of users per server and so on. I've no idea where the figures actually came from.

I've distrusted Gartner ever since.

Re:Think about this...

[Lozzer]

Yes, but any 16 year old with a cable modem and a CD burner can download a copy of any Linux distro and have a REAL, robust, scalable server OS to play with, free of charge.

They'd probably need a computer too, unless they can operate the interface on their peripherals manually somehow. Then I wouldn't put that past kids today, won't be long until they're directly wired.

Re:Slashdot should not assume its readers are MORO

[elegant7x]

Gosh, I'm down to posting a link to the Hot 100 analysis almost once a week anymore. I'll let you find it yourself this time: go to groups.google.com and search comp.os.linux.advocacy for "hot 100 uptimes", and read what you find. Or do a new analysis yourself, and post your results for us.

You do realize of course that there is a diffrence between a computer being on for a long time and it's opperating system being "W2K IS a serious Enterprise-ready scaleable and reliable OS." -- I mean, those uptimes are like 5 years. How the hell is a 2 year OS going to be able to beat that?

IDC and Gartner are counting different things

[Army No Va]

IDC was counting Linux licenses sold by distributors. Gartner is counting Linux pre-installed on new hardware, a smaller number.
Much of what IDC counts gets installed on older machines or even "client" PCs to be used as a server. Gartner is counting new server HW with Linux only. And these are annual rates, not install bases.

However, both have a hard time with accuracy due to the nature of OSS distribution methods.

Moderated 'redundant'?

[crond]

How can the first post (not counting the "I got the first post, love you mum!"-mumbo jumbo), be moderated redundant?

Tore

crond@undernet
Norwegian Linux Community

Lies, damn lies, and statistics

[connorbd]

Seems to me that the Gardner Group study may be right on the money. What's interesting about it is that even if it is in fact true, it's a pretty meaningless study for precisely the reasons Roblimo went into.

And that's the big problem with statistical studies -- surely there are plenty of Linux servers running under the table that no survey is going to find because the techies' boss doesn't even know about it. Web servers? Okay, that's publicly accessible stuff, but only a small fraction of the server market.

What I find especially interesting, though, is that as slanted and pointless as the survey is, it still works in Tux's favor over the long run, at least a little bit...

/Brian

Re:Gartner smells like Ziff Davis

[Spoing]

Damn, and me with no points. Mod that one up please!

Re:Who buys systems?

[clare-ents]

So, in my company, the box[es] that do RADIUS authentication aren't servers then?

We glued two old desktops together, installed linux + radiusd on both of them and wrote some easy admin tools. Then we deployed them. We're aiming to get automatic failover going between them too. Radius is mission critcal to us.

Why are these machines not servers, and why would it have been more cost effective for us to have bought two new machines to do this?

Re:Newton's law of slashdot posting

[gilroy]

Blockquoth the poster:

Who would have thought the tech sector would create such starry-eyed romantics (as many online activists seem to be)?

Actually, it doesn't surprise me at all. Techies, in my experience, tend to be the hopeful, look-to-the-future types (if only for the gadgets) and that wanders easily into ideological rigidity.

Re:Gartner smells like Ziff Davis

[Twisted Mind]

I think that is exactly a problem, Consumers actually *wanted* text-based DOS (GUI's are childish) and they got. After that they *wanted* no real full stable 32bit OS but demanded an simple DOS/Windows combination and they got it.
Apple isn't much better, but they at least droped compatibilty to the max for an real (not real-time :-) ) stable 32 -bit unix-based OS.

While this is not a MS-only thing, this sounds actually

Re:Gartner smells like Ziff Davis

[Twisted Mind]

Just because of the IBM-clones.

1 word

[vsync64]

Slackware.

--

Re:1 word

[Snootch]

Slackware.

Damn, you beat me to it. That said, the minimum supported with -current is 16M, which is beginning to drop my stuff off the end :-(
That said, Linux still does run well on old hardware, it's just that the definition of "old" moves with the definition of "new". Besides, I wouldn't think twice about installing Win95 on an i586 (Win98 would creak BADly...), but somehow we're averse to Slackware 4 'cause it's too old...

43rd Law of Computing:

RTFA (Read the Friendly Article)

[grue23]

* As someone else mentioned, Gartner said that they asked end users what OS they had installed on their new servers, the survey had NOTHING to do with what came pre-installed. RTFA.

* I find the Gartner numbers much easier to believe than the claim that GNU/Linux holds 25% of the server market. Gartner is well-known as an excellent objective research firm. Also, from my personal experience as a former client-facing employee of a large internet consulting firm, I know that there are a lot of companies out there still nervous about using GNU/Linux for their servers. The internet consulting firms themselves also have excellent relationships with Microsoft, Sun, and Application Server providers, so are incented financially and with training to recommend more commercial products to their clients. For smaller inexpensive sites, clients tended to lean towards Windows, and for the larger ones they tended to lean towards Solaris, at least in my experience. The one place I saw GNU/Linux getting used was at the web layer, with the RedHat/Apache combo. I never saw Linux used for an application or database server on any site my company built, and I know that it was suggested by some technical leads then turned down.

Re:RTFA (Read the Friendly Article)

[grue23]

> Gartner said that they asked end users what OS they had installed on their new servers

Even you are claiming that the survey is talking about preloaded OSes. Who installed the OS if they "had" an OS installed?

The word 'they' in my sentence refers to the END USERS. Good morning. The END USERS reported what OS they ran on their new servers AFTER they set them up.

Prediction

[WillRobinson]

Microsoft will win for a long time. They got the money to throw at programs for the masses.
But wait, the .net revolution, will cause companies to evaluate cost of ownership, data possession and security. NO ceo in their right mind, will allow their company data to be warehoused off site, or only can be used if they continue to pay every year just to use the same program to access thier data.
I could see somebody saying. "Oh im sorry, I cant get you those numbers, we didnt pay for our program this month."
I will be interested in seeing how many servers become linux servers in the .net revolution.

Think about this...

[FortKnox]

How many UNIX system admins are there in the world? What is their pay range?

How many Windows system admins are there in the world? What is their pay range?

Lets face it, my 16 year old sister can learn to be a windows admin from one of those tech schools that'll make you certified in 3 months. And you can pay 5 of those newbie admins for the price of one unix admin.

Its not that I enjoy these numbers. Honestly, I think if you want to run a windows server, you'll need the extra admins to take care of the additional crashes and glitches, so it isn't cost efficient. But John Q. Manager doesn't know that. Such is life.

Re:Think about this...

[mikethegeek]

"Lets face it, my 16 year old sister can learn to be a windows admin from one of those tech schools that'll make you certified in 3 months. And you can pay 5 of those newbie admins for the price of one unix admin."

Yes, but any 16 year old with a cable modem and a CD burner can download a copy of any Linux distro and have a REAL, robust, scalable server OS to play with, free of charge.

And Linux is getting easier and easier to use all the time.

Plus, Linux and Unix knowledge is WORTH more in the workplace than `Doze knowledge, which is a dime a dozen. I personally am considered very valueble in my department, in a top company that is investing a BILLION dollars in Linux, simply because I am a competent superuser. I'm not even good enough to be an engineer yet (but I'm working on it, and here in Durham, NC is a great place to do it).

Re:Think about this...

[maxpublic]

Add to this "do you honestly think *real* hackers use Windows? C'mon, that's the operating system for *your parents*!"

Hook, line, sinker.

Max

Re:Think about this...

[Snootch]

Lets face it, my 16 year old sister can learn to be a windows admin from one of those tech schools that'll make you certified in 3 months. And you can pay 5 of those newbie admins for the price of one unix admin.

I fear you insult your sister's intelligence! :-)

Seriously, though - ordinary end users used to use DOS, and unix is if anything easier. Three months is easily enough time to learn to be a UNIX admin, certainly to the level that you get from one of the MS classes you talk about.

43rd Law of Computing:

Who buys systems?

[wytcld]

Who buys systems when we can just assemble them from all these parts lying around?

Re:Pre-installations were not involved in the numb

[_ganja_]

I don't really now much about XP, I'm happy using Linux on the desktop BUT I am now curious about these features (#1 & 2). How is M$ going to hurt MP3 in XP, of course its no surprise given their own formats exist. Also why break CD burning software deliberately? What do they gain from that and how are they going about it?

Could be nice avocacy material.

An Alternate Study?

[Alien54]

Probably the most damning would be this customer satisfaction statistic:

Of all servers shipping with Windows software, as many as 50% (whatever the number is) have the Windows software erased and replaced by an alternate system that the owners believe to be simply superior, simple more usable and suitable (insert your favorite adjectives) for their own purposes.

Someone ought to compile this statistics and issue a white paper on the increasing number of people who erase their windows systems.

I am sure it would be damning.

Check out the Vinny the Vampire comic strip

Re:Pre-installations were not involved in the numb

[Wavicle]

Well, let's read some of the other paragraphs:

"While I accept my results may not include some desktop and workstations configured as Linux servers, I simply do not believe that Linux is shipping on 25 percent or more of all new servers and I just cannot believe the IDC figures," Hewitt said. "They are simply pushing the envelope and overstating what the operating system is actually doing."

If the majority of customers got their Linux via download or some other way, "then the market is in even worse shape than my survey shows," he said. "How many support contracts are vendors going to get from those customers? I have already told Red Hat that they are stretching the numbers they put out to the marketplace, which resulted in a very lively debate."

These paragraphs seems to imply that he was looking for what OS the machines were shipping with. And accepting that if the 25% of all servers are running Linux (i.e. the IDC study is correct), then those customers aren't buying support contracts.

The wording of the article is confused. What you quoted makes it sound as if they were looking at what OS servers bought in the past three months are running. What I quoted makes it sound like they wanted to know OS the server shipped with.

Can anyone make sense of this?

Re:RTFA! Doesn't matter what it SHIPPED with...

[ichimunki]

First of all, I think the sample is rather small. Around 750 people-- who were asked about this issue during a phone call. There is some wiggle room for error that they don't mention at all. I'm no statistician, so I can't say what the plus/minus factor is.

Second, this reports only sales of servers-- yes, I know they're asking people who have had a chance to install a new OS onto new hardware, but what percentage of the installed base of servers is actually new within the last three months? How many older servers are finding new lives or being passed along using Linux or a BSD?

What they are not taking into account, and given their sponsorship by MS, we can only assume they are not going to mention unless someone else does first, is that this doesn't mean that there aren't a lot more Linux servers out there than the 8.6% or whatever it was. This also seems to deliberately obscure the fact that you don't need a server class machine purchase to obtain a Linux server. Any new Pentium 3 with a large enough hard-drive is more than capable of handling intranet services for smaller businesses. Stuff like web-proxies, central DB, firewall, web development test system, custom programming, file server, etc. No reason to go buy a really expensive server for a lot of this stuff if you only have 2-50 employees.

The upshot of the way they present this, by trying to say that previous Linux studies were wrong when it doesn't sound to me like they measured the same thing at all, seems to me to be a strategy designed to discourage people from considering Linux seriously. And frankly, unless I'm a product manager at Red Hat or VALinux why do I even care about penetration? The three things I care about are technical quality/reliability, usability, and cost to license. The only way any of these are in danger from penetration statistics is if there is so much incompatibility between my server and the rest of the world that usability becomes a distant pipe dream. And from what I've seen Linux plays pretty nice with other OSs.

Re:RTFA! Doesn't matter what it SHIPPED with...

[ichimunki]

Don't be stupid-- did you even read my post? I never misinterpreted their report. I clearly stated why it is mostly irrelevant to those of us who are not product managers at hardware and software companies-- it measures sales, not actual penetration. Your comment about 486's in the basement makes it sound like you may be biased. Kind of like when Miller calls Linux a "niche" and Ballmer calls it a cancer. It's nothing more than peer pressure and scare tactics at the corporate level.

Re:Gartner smells like Ziff Davis

[sulli]

If it produces poor quality research on a regular basis, people will stop subscribing or purchasing its reports.

Which is precisely why I don't subscribe. Crappy research ("Paper airplane portals will be a $2B market by 2004.") Waste of money!

Re:Gartner smells like Ziff Davis

[sulli]

Find some predictions Gartner issued a few years back and compare them with the state of the world today. I suspect their accuracy will be demonstrably poor.

Amen! I bet some investors in a wide range of dot-coms and "enterprise B2B exchange portal" thingies would love to find the guy who wrote that these would be multi-billion dollar markets by now...

Re:Netcraft Statistics

[Denial of Service]

Apache doesn't equal Linux, big guy. Apache runs very nicely on actual enterprise OSs as well as your infallable Linux so your argument leads to precisely cob.

Nice try, though. I'm more surprised that you haven't been moderated to the moon for your Linux masturbation, though.

---

Re:Do we have to...[OT]

[Denial of Service]

It's funny that you have "The Great Unwashed" and RMS in the same paragraph. Maybe they can relate to each other more effectively...

---

Re:Netcraft Statistics

[Denial of Service]

Dear God, man. It's ok to point out that someone's wrong without being such an ass.

You're probably right... I often get carried away in light of the stunning lack of openmindedness around here and evidently overdo it when I get the chance. You're actually very correct, it basically does come down to money but an unfortunate number of zealots are not willing to concede that point and instead pump Linux as the second coming of Christ (with less hair), especially in comparison to Microsoft products which have an unquestionable place in the market.

Hey, I have a Linux/Apache server running three feet from me right now that is pushing 90 days of uptime, so obviously I have faith in the combination (and I can't argue with the price for a privateer like myself) but people like the one I originally replied who proudly point at the Netcraft stats as "proof" of Linux's upward swing really just don't have a sniff.

---

Another possibility

[dropdead]

did they ask of the new servers you bought were they completely new(expansion) or replacements for old servers. If your old Linux server is working and that MS box running exchange needs replacing you will see better numbers for MS.
After all how many times has it been pointed out you don't need to buy new hardware to get Linux to meet your needs.

Re:Pre-installations were not involved in the numb

[Pru]

So some.. some practices that are bad.. make the whole company evil? Thats just dumb, Microsoft is a big fucking company, they have lots of parts. What about there mice, are they evil to? What about easy of use os's.. man those are evil. What about making a os that your mom can use without 8 hours of training, thats so evil. Sure these may not be unic things, but linux is evil, its technology prohibitive from users in the inner city... get some logic.. all of MS isant evil.

This isant what you think

[Pru]

The statistics published on the top of this article arent contradictory. They are just different.

NOOOOO this isant another instance of making statistics show what you want.. this is another instance of a /. moron posting shit that is wrong.

For Instance...

Only 8.6 percent of servers shipped in 3rd quarter 2000 were running Linux, claims a recent Gartner Dataquest report. A previous study published by IDC estimated linux held about 24% of the server market share. Unsurprisingly the Gartner study was partially commissioned by Microsoft

It says 8.6% of servers SHIPPED... that is wholly different then 24% RUNNING linux, as would be in stated SERVER MARKET SHARE.

The /. employee that posted this errored just as much.. damn they are just as dumb as the trolls sometimes.

Re:This isant what you think

[geomcbay]

I've come to the conclusion that the Slashdot editors are in fact not trolls, but ingenious page-hit-count-generators. Some of the posts they allow are so outrageously stupid they HAVE to know its flamebait. My guess is they do this because the flamey articles tend to get a lot more page hits, posts, reponses, etc than the stuff that makes sense.

Oh well, all the page views in the universe isn't going to help VA Linux survive through this year.

Re:This isant what you think

[geomcbay]

I agree. Losing SourceForge, in particular, is a big deal. But I don't see how they will manage to survive.

Hopefully people who host their code at sourceforge will at least start making copious local backups in the event the site just disappears one day, along with VA Linux as a corporate entity.

Re:This isant what you think

[pherthyl]

Loosing VA Linux would be pretty harsh.. Not that I give a rats ass about them or their products but what about all the sites that they host? You have to admit that the sites they host(slashdot, OSDN, sourceforge(i think those are all VA, correct me if I'm wrong)) are really useful to the whole open source community.
Even from an entirely selfish viewpoint, it would be a terrible shame to loose such a valuable resource.

Re:Lets talk about "recycled" servers

[Jaysyn]

well your damn sure not going to throw ms-dos on a Pentium 60 and call it a file server..

Jaysyn

Re:We need a FUD log

[Jaysyn]

MicroDot????

Jaysyn

RTFA! Doesn't matter what it SHIPPED with...

[ColdGrits]

As they state quite clearly in the article, "We went to end users, rather than looking at just sales numbers, and asked them what servers they had bought over the past three months and what operating system they had installed on it over the same period" (my emphasis).

See what they state? They asked what OSes people INSTALLED on the systems they had bought, NOT what OS the system shipped with.

--

Re:RTFA! Doesn't matter what it SHIPPED with...

[ColdGrits]

OK, which part of "A recently released Gartner Dataquest report, sponsored by several companies including Microsoft, found that just 8.6 percent of server shipments in the U.S. during the third quarter of 2000 were Linux-based" didn't you understand?

The report made it perfectly clear that they were comparing ONLY servers (for ALL OSes) that were biught within the past 3 months, period.

Now, if you choose to interpret that as them claiming only 8.6% of ALL server are Linux-based, then that is your problem, not theirs.

Yes, there are a great many existing Linux servers out there, just as there are a great many NT, W2k, Solaris, QNX, *BSD and goodness-knows all sorts of servers out there. That, however, was not what the survey was designed to find out. As it made clear, it was looking to see what %age of NEW servers were being installed with which OSes, and that is what it did.

I guess if I were a manufacturer of new hardware, I'd be interested to know what OSes people buying new equipment wanted to run - I'd be far less interested in what OS someone was running on their old 486-in-the-basement...

--

Hmmmm

[friday2k]

Does anybody have numbers for things that really count? Things like how many of the real heavy traffic sites are run on Win2k vs. Linux vs. *BSD vs. HP vs. IBM? For many of the big companies stability counts more than free. Win2k is maybe not doing good here, but does _free_ Linux do good here (vs. e.g. HP or IBM) or are commercial Linux implementations through IBM (and that can be AIX or Linux or OS/390 for the real insane)? It shouldn't concern the community if more servers are running Linux or Windows or Amiga OS but it should more count what people are doing with it. And I would not bitch too much about Win2K, there _are_ big sites that are running on it. And you _can_ make it secure. And you _can_ make it performant. Not necessarily if the admin is a moron, but ...

Re:Hmmmm

[jsse]

Well, actually after economics turmoil corporations are seeking alternatives in the implmentation of enterprise systems.

E.g. my company listed Linux in the evaluation process when starting new project, which is so amazing - our company never ever spend any resource on R&D and every adoption must be known technology and justified(read: have big corp. behind). The move is like seeing a dinosaur flying to us.

Our internal papers even positioned Linux as the midway between Win2k and commercial UNIX. It might not due to the increase in awareness of management, it might due to the budget consideration.

They solved budget problem, we are happy with Linux. It's a win-win situation for us.

(That's only one specific case for our corp., I didn't mean to generalize - in case you argue)

Re:Pre-installations were not involved in the numb

[boomi]

that's why he said "in the Windows world".
For MS, third-party software is bad because it lowers the use of their software, just ask Borland how difficult it is to release developer tools for windows...

Re:Those in the middle

[Technician]

Either dual boot or have machines of all flavors. Windows ME/2000/98/95/31/CE/Pocket, Mac, OS/2, PC DOS, MS DOS, DR DOS, UNIX, PICK, Netware, ... It's too much to learn all of them well enough to be proficient in them all, so they tend to favor one over the other for whatever reason. Usualy the reasons has to do with compatibility, ease of use, application requirements, and support.

Is having MORE servers with your OS really better?

[kstumpf]

It takes more Windows servers to accomplish the same thing fewer Linux servers can do. So if I only need two Linux machines to do what five Windows servers would handle, does that mean MS has more market share? Well, of course it does! They sell you overpriced, underpowered, unstable bloatware.

Are corporations buying more Windows servers than Linux servers? Yes. Are corporations buying more Windows licenses than using free software? Yes. Are corporations buying more MS support contracts than Linux support? Yes. Do the majority of corporations operate efficiently? Hell no. Are corporate decision-makers aware of the benefits of open source? Typically not.

Studies like this do nothing to prove or disprove the value of opensource. What I would like to see are comparisons of similar-sized companies that use either Windows or opensource... how do their server-farms compare? Who's more stable, more secure? Who's budget is lower? How much does each company spend on support, hardware, etc? How about some real side-by-side comparisons of real-life scenarios, rather than a guess of how many servers are out there?

Another thing - all of my servers were bought without Linux (some with no OS, some came with windows). I download distros and keep them on an FTP on my LAN. I install from that with the bootnet image. Even if Gartner asked the purchaser what the machines were intended to run, they would not have known and said windows since we are main an MS shop, except for my systems.

using the old hardware.

[saintlupus]

for my older machines -- 486s, 586s, moderate amounts of RAM -- that I have to track down pre-2000 distributions, or settle for crippled distributions such as SmallLinux (very cool, can do X-Windows in 4 megs of RAM, but also states very clearly "this is not a complete distribution"). I think our convincing arguements about how good Linux is with old machines are becoming less convincing as older distributions become more obscured or drop off the Web entirely.

*cough*

running the newest official release on a 486. nary a hiccup. of course, it doesn't come with a gig and a half of useless preinstalled cruft, which really can't hurt.

--saint
----

We need a FUD log

[wrinkledshirt]

I suppose some would consider Slashdot to be its own library of MS public relations exposure, but I really think we really need to make a chronicle of all the ways that Microsoft has handled the Linux OS.

Methods they have used so far:

• Mockery (Ballmer)
• Misdirection (this study)
• Defamation (Alchinn)
• Conspiracy (Hallowe'en papers)
• Divide and Conquer (recent attack on GPL)
• Rhetoric (that "truth about linux" article they had at microsoft.com for a while)
• Hostile advertising (ads over in Europe)
• Sending subliminal messages to morons who happen to moonlight as ZDNet editors
• Hypocrisy ("How can we be a monopoly with this wonderful Linux thing going on?")

We should compile all this stuff and put it in a central location so that we can refer to it at a later time. I bet it wouldn't take long for all the inconsistencies in their arguments to fall apart, and it would make for great debate fodder.

Not very scientific....

[StarTux]

None of these studies could stand up to any scientific scrutiny....

1. The pool of so called "professionals" is very small, only 724 (from memory...). This hardly represents anything.

2. What sort of questiosn did they ask to gauge the competence of the end user? If its partially sponsered by MSFT, then did MSFT get to do the screening?

4. I have noticed that people with Unix backgrounds go the *nix route, Windows background favour Windows a little more...

5. All the statistics are meaningless and unscientific as most of his points are opinionated. Just give the facts, not the babble.

6. If Redhat had comissioned it it could easily have been 8% of shipments were Windows based, amazing how easy it would have been for them to pay for the result they intended...

7. Only time will tell who is going to do the best.....

Hope that made sense.

Re:Slashdot should not assume its readers are MORO

[arfy]

Just a slight note: I hope that when you send reports to your clients, you say "The other managers and I" rather than "Me and the other managers at the consultancy where I work"...at least when you're starting a sentence.

Me read reports too, but me try write better.

Also: "When my corporation pays for research, we absolutely do not want them to tell us what we want to hear. We want to know the FACTS. Bill Gates (the world's most succesful businessman) is no different."

You're assuming Microsoft wants facts and not marketing crap to try to prop up W2K. You know, stuff marketing can put out alongside statements like "W2K IS a serious Enterprise-ready scaleable and reliable OS." Which might be true in comparison to LAN Manager, but not if it's up against UNIX.

Finally: "Just because Microsoft commissioned the research does not render it invalid."

Agreed. It does, however, render it highly suspect. That it's from Gartner renders it likely invalid. They haven't been trusted by any of the techies I know since about 1996 and even the sales types I know don't take them seriously anymore.

Bullshit.

[Magnus Pym]

The Garter group survives by some of the most flagrantly unethical (and in my opinion, criminal) tactics ever displayed even in an industry where such things are common. As for driving Garter out of business if you do better research... what utter nonsense! That is like saying that you will drive MS out of business if you produce a better OS. In an ideal world, maybe. Not in a world driven by greed and corruption where manipulating and managing the public perception of things has been reduced to a science.

Magnus.

Dur Hay

[Apreche]

As has been said many times before on Slashdot. Most servers that you buy come with Windows installed. Most computers you buy come with windows installed. However smart IT departments wipe the drives on all their servers as soon as they get them and change to some *nix. Sure some of them keep the Windows 2000 Server, but most don't.

Most servers are sold without a free operating system. Most are wiped. There's no way of getting the stats on that without a large survey of IT departments. That's the truth right there.

Re:Dur Hay

[xandria]

Hey dummy, read the Gartner piece. It specifically said that the respondents were asked what OS was (or will be) running on the purchased servers, not what OS was shipped with their server. Gartner's Hewitt said --

"We went to end users, rather than looking at just sales numbers, and asked them what servers they had bought over the past three months and what operating system they had installed on it over the same period," Hewitt said. "There was no question about whether Linux was preinstalled or not, we simply asked about new shipments and this is what we found."

You assume too much

[TheLinuxWarrior]

I'm not trying to bust your balls too hard, but do keep in mind that all of the major hardware vendors are shipping systems with Linux on them. My last company bought several Dell servers with Redhat preinstalled.

Re:one other word

[Anthony Boyd]

Well thanks for that bit of info. I assume that implies a GUI in a very small RAM space -- if it's running well on a 386, I suspect it's able to do X and blackbox in perhaps 4 megs of RAM? I could go check their site, but I'm very tired of checking *nix sites. Besides, I suspect I will be the only to read this, considering how long ago the OP was made.

In any case, my complaint was more about Linux in particular than Unix in general. I know the BSDs are more refined (especially OpenBSD, now there is system security done right), but I've shied away from the BSDs for two reasons. First, I'm a Web developer, I want to concentrate on building Web pages. All I need is a text editor and a browser. So I don't want to geek out with the box itself -- I'm not a sysadmin and don't want to become one. And I just haven't found the rpm/deb friendliness on BSD. Second -- and I could probably only get away with saying this in a dead thread like this one -- but I really feel in my gut that I don't like the BSD community. This is not a community of people looking to grow. They're trigger-happy with the "rtfm" crap.

I know it doesn't have to be that way. Back when I played with Perl 4 in 1995, the community was quite friendly, but over the years got really exhausted (as a whole) by the influx of new users. Eventually their FAQs and massive, sprawling man-pages were used as weapons to shut up anyone with nearly any question. The Perl community stagnated, and started losing out to technologies like PHP, which has an absurdly friendly community (see phpbuilder.com). In recent weeks, the Perl community has begun to change -- starting a newbie mailing list (see perl.com) and a general trend toward new-user outreach. To see a mature technology and entrenched user-base deliberately decide to reasses and reengineer their group-mindset is an impressive thing. It gives me hope for BSD, but I don't see anyone in the BSD community following Perl's lead. And in the Open Source world -- and much of the point of the Cathedral and the Bazaar -- is that it's all about community.

Always enjoyed your sig, by the way.

Re:Lets talk about "recycled" servers

[Anthony Boyd]

Dude, I don't know anyone that "takes that old machine in the corner" and loads Win2K on it ;)

Actually, I'm having trouble taking "that old machine in the corner" and loading Linux on it. Firsthand experience over the past month of reading, experimenting, and downloading distros has revealed that most 2000 and 2001 Linux distributions really don't work that well for old machines anymore. You can see it with Mandrake and others optimizing for 586 or even 686 processors now. You can see it in Debian 2.1 vs. 2.2 -- the memory requirements for 2.1 (1999) are 4 megs ram minimum (16 recommended), for 2.2 (2001) the minimum is 12 (64 recommended).

What I've found is for my older machines -- 486s, 586s, moderate amounts of RAM -- that I have to track down pre-2000 distributions, or settle for crippled distributions such as SmallLinux (very cool, can do X-Windows in 4 megs of RAM, but also states very clearly "this is not a complete distribution"). I think our convincing arguements about how good Linux is with old machines are becoming less convincing as older distributions become more obscured or drop off the Web entirely.

Re:Is having MORE servers with your OS really bett

[pkesel]

When you're selling, all you care about is how much they're buying. High ideals about which is better technically don't keep the stock holders grinning. They don't care about stable or secure. They care about how many boxes went out and how many checks came in.

Links.

[V50]

Here are some relevent links:

An Opinion from OsOpinion
The Gartner Report Itself

--Volrath50

It serves it's purpose.

[rohar]

What I've noticed in my little world, being employed at a Canadian GIS consulting company that mostly works for urban municipalities and government, is that middle managers choose Microsoft because it is safe.

A report like this, whether it actually reflects a true market or not will be used by the millions of middle managers to back up their decision to choose Microsoft.

I worked on a web application project for a provincial crown corporation, and the entire web infrastructure was built on MS products (NT,IIS,MTS,SQL Server), even though the legacy system was built around Sybase 11.3 on AIX, and there was a ton of Unix experience around the building, they went with M$ products.

Turned out that MTS couldn't do a transaction against Sybase, IIS didn't scale at all, etc. etc.
That is not the point though. No one was going to make the decision to use something like GnuLinux/Apache/PostgreSQL in any Canadian government organization, because the middle managers aren't technical enough to take the chance.

The factor in this case is IBM. The only way GnuLinux is going to get into these organizations is if IBM, HP, or Sun puts it there.




Whatcha doooo with those rollin' papers?
Make doooooobieees?

Re:Gartner smells like Ziff Davis

[jsse]

I'm not the one you bashed, but I was wondering where do you get this information?

As has already been stated, Gartner asked end-users what they installed on their computers after they bought them.

As one of a Gartner customers, we failed to get hold of the detail information on how they did the survey, they (inofficially) said it's commercial secret. All we know is the market segment and sample size. So, are you sure they really did?

Garter is a research agency that has its value locked in its reputation. If it produces poor quality research on a regular basis, people will stop subscribing or purchasing its reports.

Good point, but in fact customers choose Gartner because they want to get what they want to hear, and Gartner chooses who to list in their report. Once my friend's company complaint Gartner on their unfair comparison to their product. Their response was like "You were lucky we ever listed you" attitude. Gartner offered to come over to evaluate the situation, turn out gave them an opportunity to make business. My friend has then grown up thereafter any never take their study seriously.

Are you arrogant enough to believe they haven't thought of this?

Are you naive enough to believe they haven't thought of doing study in favour of big corps. for long term benefit?

Garter is not in the business of assessing obscure technical facts. They provide a strategic business perspective on technology. Tech-heads are not their market. They don't care about the operational aspects of the technology. The people who run business (which, most commonly, are not tech-heads and have different skills) are their market. They care about the strategic implications of the technology and longer-term market trends.

Hmm, isn't that exactly the major problem here?.....no wonder why they rated Rambus having brilliant future....

But, the village idiots on slashdot are always willing to provide legal advice, assume everyone in management is a PHB, that companies never ever know what they're doing, that everything is part of a conspiracy and that anyone who doesn't know how to write a sound compression script using bash is an idiot.

village idiots on slashdot? Compare to people like you who are so naive to believe that reports from big and rich companies must always be trusted, we are real idiots.

In summary, read the bloody report and get some perspective before spouting off.

I read those bloody reports quite frequently for my job. Have you ever wonder how do they come up with those probability factors that predict future trends on something, and how accurate do they turn out to be? I'd be much grateful if you could talk Gartner's to release the details, formula and source data of their research. (No Sir, it's company's secret!)

Re:Netcraft Statistics

[Demerara]

"Especially considering third world countries, who are by definition poor, are still striving to get online."

Not just online but just get computing at all! The cost of hardware, software and bandwidth is very high in developing countries. For example: T1 in Guyana approximately US$26,000 per month. For those of you running at 1600x1200 that's US twenty six thousand per month.

Anything at all which saves money will be taken up - especially as the pressure to respect intellectual property rights (aka Government departments actually having to buy software - "you mean I have to pay US2,000 for that CD-ROM?")

Having said all that, Guyana is still pretty much an MS shop. Our foreign minister (since "laterally promoted") even invited Bill Gates to come and visit. I think they wanted him to buy the country or something.

But I digress. Linux is slowly making it's presence felt here. And, natch, almost ALWAYS installed on hard disks recently fdisked free of Windows!

Re:Pre-installations were not involved in the numb

[mikethegeek]

" Well thats normal /. dumbfuckness.. everything AOL or MS is evil.. they could donate $1 Billion to linux.com and they would be strewn as evil monopolists. STUPID FUCKING SLASHDOT TROLLS"

Really? As I recall, most /.'ers like Mozilla, which is a Netscape derivitave, which has developers that AOL pays...

That is a good thing AOL does, for sure, if for no other reason than to help Mozilla is to hurt M$. Remember, the enemy of your enemy is your friend...

As for Microsoft, I make no apologies. They ARE evil. Just look at what `Doze XP is going to do to people who think they are buying the greatest thing since MS Sliced Bread(tm).

XP:

1. Deliberately hurts Mp3
2. Breaks CD Burner software
3. Prevents you from upgrading your computer in an unlimited fashion without calling M$.
4. REQUIRES activation (and giving out info) to be a legal license.
5. Is a very MINOR technical upgrade from Windows 2000, which given what XP is DOING to you in #1-4, makes it a negative return. 2000 is by far the better OS.
6. Has a GUI that would make anyone outside Miss Shirley's Romper Room throw up. I much prefer KDE 2.1.

Not only that, but MS is sending out goons to call the GPL license "Un American" and to lobby for what cannot be stated as anything but the right to "embrace and extend" ANYONE'S code.

Re:Gartner smells like Ziff Davis

[mikethegeek]

"As has already been stated, Gartner asked end-users what they installed on their computers after they bought them. Not what was pre-installed on their computer. Implication: either Linux doesn't have the marketshare zealots want to belive, or accurately assessing server marketshare is difficult. You decide."

NOT true. Here's a quote:

""The study results indicated that in the traditional server market in the United States during the third quarter of 2000, 8.6 percent of server shipments were Linux-based systems."

So they did base it on SHIPMENTS. I'm willing to bet that most GNU/Linux servers ship with Windows, then are replaced with Linux. Also, this does not count hand built servers, of which I'd bet a large percentage are Linux servers.

As I said in my previous post, this "method" of determining how popular an OS is used by Gartner isn't any more scientific than determining the most popular video game by what cart/CD came with the system.

Gartner smells like Ziff Davis

[mikethegeek]

In my previous job, my company did a lot of contract work for the State of West Virginia, which uses the "Gartner Group's Tier" standards for all their purchases.

Many things on it were illogical, such as the blatant RAMBUS memory cheerleading, etc, despite benchmarks that proved that PC-133 SDRAM was faster on Pentium 3 systems than RAMBUS (and was cheaper).

Basically Gartner, like any company of it's type, says what it's paid to say. Why do they do this? Because they wanted to get paid. Remember when Ziff-Davis would only talk about Liuux when parroting Microsoft's "Haloween Documents"? That changed overnight when Linux companies formed and started advertising in ZD magazines.

The same will happen with Gartner when a Linux company buys their seal of approval.

As the article stated, this "study" used the flawed model of judging server OS's by what OS shipped with the system, rather than actually surveying what was actually USED on that server sold.

This is like measuring what radio station is #1 in the market by what station the radio was set to when it was sold. Or how many people buy Ashland Gasoline because that was the fuel in the car when it was sold...

Unfortunately, Gartner's flaws are not well known by the people who use them as the HOLY WRIT (usually the non-techical people in corporate and government bureaucracy).

Re:Gartner smells like Ziff Davis

[NutscrapeSucks]

IDG's not any better -- there's been complaints over the years about how they tally the sales of Mac software for example. But at least there's not common allegations that IDG is dirty as their are with Gartner (remember when Gartner published two reports at the same time, one reading "Linux is great (0.7 probability)" and the other reading "Linux sucks (0.7 probability)"?)

I'm sure that the IDG numbers are total fluff -- they probably took total Linux downloads and ass-umed that X% were going to new server installs, and worked backwards from there. The fact that the 24% number is "shipped" is especially bogus because most servers don't ship with an OS at all.

Re:Gartner smells like Ziff Davis

[james(honest)]

The relavent quotes:

"We went to end users, rather than looking at just sales numbers, and asked them what servers they had bought over the past three months and what operating system they had installed on it over the same period," Hewitt said. "There was no question about whether Linux was preinstalled or not, we simply asked about new shipments and this is what we found."

"While I accept my results may not include some desktop and workstations configured as Linux servers, I simply do not believe that Linux is shipping on 25 percent or more of all new servers and I just cannot believe the IDC figures," Hewitt said. "They are simply pushing the envelope and overstating what the operating system is actually doing."

A recently released Gartner Dataquest report, sponsored by several companies including Microsoft, found that just 8.6 percent of server shipments in the U.S. during the third quarter of 2000 were Linux-based.

For those of you (cafeman) not familiar with the use of spin, let me point out how the first paragraph seems to indicate that the figures in the second paragraphs are what Microsoft want you to believe (i.e that only 8.6% of servers are actually running linux). They are clear about how they interviewed people about their usage.

However, note that they are repeatedly very very sepecific about the word shipment. By describing their interviewing process, they lead you to beleive that these usage results are the ones they are referring to when they say 8.6%.

In fact, it is perfectly possible that they have two sets of numbers and know full well that 25% of servers are running linux and at the same time state, correctly, that only 8.6% of systems are shipped: "I simply do not believe that Linux is shipping on 25 percent".

So, given that a) they are paid by Microsoft and b) they are very deliberate in the choice of misleading words while at the same time specifically avoiding saying "only 8.6% are running linux", I'd be inclined to say thats its the usual manufactured statistics.

Jamie

Re:Gartner smells like Ziff Davis

[Snootch]

A company can succeed by raping customers for a limited length of time. Eventually it has to produce a product people actually want, or go away. IBM knows this all too well. Microsoft will learn it too.

Boy how I wish that were the case. No, actually, it is. THe problem is that with effective marketing you can make your product what the customers want. It ain't pretty, but it's true...

43rd Law of Computing:

Re:This isant what you think (CORRECTION)

[geomcbay]

Should have previewed that. Doh.

I meant to say:

I've come to the conclusion that the Slashdot editors are in fact not stupid, but ingenious page-hit-count-generators.

Which in fact makes them trolls

Re:Slashdot should not assume its readers are MORO

[geomcbay]

Heh, support for XML in the kernel.

Heh. Considering the rest of his post seemed non-trollish and mostly intelligent, I'm guessing that what he meant was support for XML in a standard way. It would be nice if Linux (*NIX in general) had one single standard library based on XML that was used for all *.conf type files. It really can be a pain editing different file types and doing things differently depending upon if the software you are configuring uses ':' to seperate keys from values, or ',' or something even more obscure.

Why you would want that support in the kernel, though.. well, its a bad idea. Perhaps the original poster assumes this will trojan horse the XML support into all distributions and then software programmers will be more likely to use the library...

If nothing else...

[Scoria]

... at least we don't have to worry about *BSD is dying troll posts in the comments for this story...

Re:Pre-installations were not involved in the numb

[astr0boy]

1. Deliberately hurts Mp3
unless you use 3rd party software

4. REQUIRES activation (and giving out info) to be a legal license.
it does require activation, but that isn't a problem if you actally buy the software. It does not give out your info, all it does is generate a unique identification number based on your hardware which gives them no information.

5. Is a very MINOR technical upgrade from Windows 2000
except most of the sales will be to the home user (via the "personal" version of XP). If you are running win2k and consider this a minor technical upgrade here is an idea: dont upgrade.

6. Has a GUI that would make anyone outside Miss Shirley's Romper Room throw up. I much prefer KDE 2.1.
hmmm... i dont seem to be throwing up... and i happen to like it. if you prefer KDE 2.1, use it and stop complaining about things that you won't use anyway, or use the "windows classic" theme.

oh yeah, i burned a cd while typing this (using 3rd party software (musicmathc jukebox) in XP).

-----

Re:Pre-installations were not involved in the numb

[astr0boy]

well winamp and musicmatch both work. i don't know anyone who isn't a compter nerd that uses other software. so if yes microsoft lets its largest comptetitors operate and if no, well, oh well...

-----

Re:Pre-installations were not involved in the numb

[astr0boy]

but isn't 3rd party non-included free software what all you linux folk want? i have never had musicmatch jukebox or winamp crash or cause other problems... infact i havn't had to restart since i installed beta 2 (more than a month ago).

-----

the real world

[geekforhire]

Who cares how many servers 'shipped' with a free OS. All I do anymore is *replace* NT with BSD or Linux.

Re:the real world

[geekforhire]

The 'agenda' I have is doing what is best for my client. A lot of the time that involves BSD or Linux....but not always. I dont care what OS a server is using as long as it gets the job done.

Re:the real world

[xandria]

Hey dummy, read the Gartner piece. It specifically said that the respondents were asked what OS was (or will be) running on the purchased servers, not what OS was shipped with their server. Gartner's Hewitt said,

"We went to end users, rather than looking at just sales numbers, and asked them what servers they had bought over the past three months and what operating system they had installed on it over the same period," Hewitt said. "There was no question about whether Linux was preinstalled or not, we simply asked about new shipments and this is what we found."

Re:Do we have to...[OT]

[fishbot]

We don't _have_ to go down RMS' road of brainwashing, but the simple fact is this: you take out the GNU tools and you will have exactly this: a kernel and a load of user software with little or nothing to hold it all together.

Linux is a kernel. GNU is an OS. GNU/Linux is my favourite way to run a computer. Thats it!

The figures are right, but are they?

[fishbot]

After reading many posts, some pro-linux, some pro-MS, some just waiting to be flamed, I think I ought to add this little snippet.

The numbers involved here are most likely completely accurate. The problem here is the reaction of people TO those figures. Sure, 8.6% of servers may be sold with GNU/Linux on board, but as so many people pointed out, it is very (although decreasingly) hard to get a machine with Linux pre-installed.

So 24% of server run Linux? Possibly, but it's an estimate. The fact is that all these figures are subject to three facts;

1) All PC's are bought with OS and preinstalled
2) All PC's are bought in their final configuration
3) The user has a gut reaction to any word in the subject matter (Linux is bad... AAAGH! Windows is bad... YAY! etc. etc.)

I entirely believe that only 8.6% of systems are sold with GNU/Linux preinstalled. I equally believe that more systems are sold as a large pile of component parts than complete systems. I know for a fact that at the company I work for, precisely two machines have been bought complete. All the rest were built in house (that includes the desktops). So how do we rate? Well, 100% of out purchased systems ran Win95 (both of them). They don't know, but they did when new.

All those who would post their immediate gut reaction should hold back and think 'do we really need to defend this? or are we just responding to the MS publicity machine?'

I think you'll find that the latter is more often the case.

the friendliness of BSD .ne. BSD communities

[teambpsi]

FreeBSD in particular installs are so incredibly awesome -- if you have the bandwidth, get the boot images and just install it from the internet.

If not, and NFS mount on LAN machine is bliss :)

And package installs from the "PORTs tree" is just too cool

I have to agree that the *BSD crowd is a bit snobbish -- maybe even elitist at times -- but the more i compare linux to bsd -- the more i'm impressed with BSD's stability and resource utilization -- but with linux its just a matter of time

Clearly the difference stems from the "one source" for the kernel -- and maybe i'm just too stuck on rebuilding the kernel -- the config files are just too easy to run through to flip things on and off -- my current kernel is parred down to about 1.4mb

As a software developer/consultant, I really like the BSD license -- I can weild a big sword on a consulting projects for companies that want to build blackboxes (network toasters et al) and don't want to give it all away.

Sure you can do this with Linux -- but when it came to selling them on a "free" OS -- all i had to do was point to the Walnut Creek FTP server information and they were sold.

I still think that linux + gnome is a much happier desktop environment then FREEBSD, but I was delighted to see it was added to the official support releases, and SUN is getting in on it too (of course this is the same company that was DEFINITELY going OpenSTEP a few years back ;)

one other word

[teambpsi]

FreeBSD...

still runs VERY happy on a little 386/16mhz

of course *I'm* not happy with it ;)

Lets talk about "recycled" servers

[teambpsi]

Dude, I don't know anyone that "takes that old machine in the corner" and loads Win2K on it ;)

Linux/GNU/FREEBSD is doing more to help "reduce/reuse/recycle"

Hey, this report isn't all bad...

[zoid.com]

Look at this statement: "A recently released Gartner Dataquest report, sponsored by several companies including Microsoft, found that just 8.6 percent of server shipments in the U.S. during the third quarter of 2000 were Linux-based. " Five years ago Linux was doing good to get notice from any big press, now Microsoft is paying for the free Linux press :)

Re:Do we have to...[OT]

[Eryq]

A bit off topic but the /. article calls it GNU/Linux. Do we really have to go down rms's path of brainwashing?

Personally, I prefer Lignux: pronounced with a silent "g", and occasionally spelled with an invisible "g".

I don't think it's a bad thing for the Great Unwashed to see those two names together. This article was about the underrepresentation of Linux in statistics... RMS might argue that GNU is similarly "underexposed", especially given the importance of GNU software to most Linux systems.

Besides, have the "GNU" out there also gets people asking "What's GNU"? To which I can respond, "Nothing much; what's gnu with you?" And then, I can tell them all about Open Source software and why it's so damned important.

Newton's law of slashdot posting

[screwballicus]

For every anti-microsoft posting, there will be an equal and opposite pro-microsoft posting.

No doubt, within minutes, the board will be awash in criticisms of Slashdot's anti-microsoft stance and defences of that stance.

I've noticed a trend with regard to this process. No matter how passionately the stoical marketeers defend Microsoft, nor how predictably Linux' yes-men define the particular news story as the turning point in the eternal battle between the forces of freedom and the forces of evil, the truth will lie somewhere in the middle. What's depressing is that those who take the middle ground are morbidly few. Who would have thought the tech sector would create such starry-eyed romantics (as many online activists seem to be)?

Do we have to...[OT]

[hondo77]

A bit off topic but the /. article calls it GNU/Linux. Do we really have to go down rms's path of brainwashing?

Re:Do we have to...[OT]

[Snootch]

I think it's important that the GNU/free software thing gets a greater airing - after all, it's what makes Linux what it is. Also, if we don't push the whole "free software" thing, we become vulnerable - "shared source", anyone? RMS is quite often correct...

43rd Law of Computing:

Microsoft responds

[DougMiller]

A couple of things I would like to point out about this study, on behalf of Microsoft. a) Yes - Microsoft helped fund this study - along with a number of other companies. The goal of the study was to get an accurate picture of how all OSes are used across a wide selection of customer types. We fund a lot of market research in order to better understand how customers use software and services. We are interested in real market data - there is no other agenda here. b) This study is based on surveys with end users in which they are asked which OS they are using on a new server that is being put into production. Where the customer acquired their OS is not a factor, so the customer could have bought a commercial OS, downloaded the OS for free, bought one copy and deployed it on a number of machines etc. There is a huge difference between these numbers and numbers reported from IDC on server software sales. I would assume, that since Linux can cost very little, there are a lot of folks who "buy" Linux but don't necessarily install or use it. Counting OS "sales" when you have an OS like Linux that can be downloaded anonymously for free or installed many times from a single copy seems a little irrelevant. The interesting data is who is really using Linux. That is the data that the Gartner survey went after. Doug Miller Microsoft Corp.

Re:Microsoft responds

[Dutchie]

Doug, learn the golden /. rule man! If you want to be heard, post EARLY! You don't earn karma points just by being a Microsoft director y'know! And now nobody but poor old bored me is going to read Microsoft's response!

Drats, I wish I had some moderation points so I could mod you up a bit and make you look uhhh 'insightfull' ! uhhhh.... uhhhh.. ok, I didn't mean that.

Ok, lemme ask you something so I don't look like a complete halfwit. Since Gartner has apparently in all honesty told Microsoft that this whole Linux thing is just a hoax, does this mean that Microsoft is now going to close the FUD gates, (hey I said Gates!) and stop that sillyness already? Leave us alone, we don't want your company's competition, software, FUD and what all, we just want to make a COOL operating system that works, and guess what, we want it because we were frustrated with what we had, not because what we had was so great, so in a way Microsoft dug it's own uhh whatchamacallit.. gate!

Now get off of my messageboard :P~

• Imagination is more important than knowledge.

Re:Microsoft responds

[isotop]

Yes - Microsoft helped fund this study - along with a number of other companies. The goal of the study was to get an accurate picture of how all OSes
Why would MS be interested in an accurate picture, when FUD is more efficient ?

not to be taken without a grain of salt...

[pherthyl]

first of all, the fact that Microsoft (partially) sponsored this "study" invalidates it completely. The public loves statistics, they're a neat, simple method of analysing that even the average moron can understand. However, most things are too complex for a simple statistic to explain. There are inevitably other factors that come in to play that can be handily ignored in order to make statistics lie. (ie, most linux machines aren't bought with linux preinstalled) And this is MS's newest weapon in their FUD war against anything that has the potential to hurt their sales. (or as they like to call it "curing the cancer") ;) Good luck MS!

Re:not to be taken without a grain of salt...

[xandria]

Hey dummy, read the Gartner piece. It specifically said that the respondents were asked what OS was (or will be) running on the purchased servers, not what OS was shipped with their server. Gartner's Hewitt said:

"We went to end users, rather than looking at just sales numbers, and asked them what servers they had bought over the past three months and what operating system they had installed on it over the same period," Hewitt said. "There was no question about whether Linux was preinstalled or not, we simply asked about new shipments and this is what we found."

To be quite honest...

[Dutchie]

I have wondered this time after time after time again... WHY should GNU/Linux users care what market share GNU/Linux has and whether some BS company like Microsoft lies about it or not? Are we not missing the whole point here that GNU/Linux is all about (well, okay, that's open for interpretation). Did Linus ever set out to write an Operating System that would actually challenge Microsoft's 'OS' ? I don't think so. Did RMS ever set out to create an anarchy that would overthrow the establishment? I don't th... ehr, well maybe :P

I would agree that 'what GNU/Linux is all about' is probably something that changes over time, but did GNU/Linux get to where it is right now by actually TRYING to beat the crap out of M$ or any other software company? I'm sure M$ would be happy to make us all believe that and to lure us to the Dark Side that way and try to make us behave like they do so we all become more predictable to them and therefore easier to beat. I don't think that some healthy competition is harmfull at all, but look at some of the people on here, I mean... it's frightening to see sometimes. Think about the 'bazaar' people, Free Software is written by the people, for the people, and not to beat the crap out of M$.

What we have here, what Free Software writers have founded, COULD IMHO very well be the foundation of a new kind of world where people will actually cooperate together to guarantee the best possible result. Getting angry with some FUD spreading CRAP #(#$)($)(^ deadbeat MORON company like M$ (sorry, had to do that ;) is really futile in that perspective.

Just deal with it, Microsoft HAS already been overtaken by the Free Software movement and they are far too late to try and undo it. They cannot stop what has been set in motion anymore. It serves no real practical purpose to worry about their spasms anymore, they'll have those spasms for a few more years until they're dead while Free Software will be thriving and enabling us all to focus on somewhat more important issues than buggy software. Just ignore those spoilt rich brats in Redmond, that's really all you need to do.

• Imagination is more important than knowledge.

I want one of those UPSes!

[blang]

Only 8.6 percent of servers shipped in 3rd quarter 2000 were running Linux, claims a recent Gartner Dataquest report. A previous study

My server, which came with win95 in 1998, was not running anything when it shipped. I had to turn on the power. Then, after a few days, it ran Red Hat, and the last year it's been running debian.

Unfortunately

[SilentChris]

Unfortunately, most Linux users "don't care what OS is pre-installed on their machines". Keyword: unfortunately. Because if that's where the number is coming from (which is perfectly reasonable to assume, and personally reasonable to continue to assume will happen) then Linux fans should demand their OS on their Dell and IBM desktop machines (yes, these configurations exist).

As for me, I could really care less what the numbers say. I'm not a Linux zealot, but I do use it half the time on my laptop (the other half of the time is spent in Windows 2000). Both serve my needs. Of the two remaining machines in the house, one runs Windows 2000 (for the kids' games -- they tried Linux and didn't like the lack of games) the other is a Mac running Mac OS X (which is a fun experiment but rather slow).

The thing is...

[Snootch]

Some of the posts they allow are so outrageously stupid they HAVE to know its flamebait

The only reason it's flamebait is because it's wrong. Unfortunately, large surveys like this are what far too many suits read, so you gotta make a noise, otherwise ppl will think it's accurate research.

43rd Law of Computing:

gartner is very far from being neutral

[allthenicksaretaken]

its pretty much common knowledge that gartner are far from being a neutral voice - despite the fact that they pretend to represent themselves as exactly that. they have a number of unsubstanted bias towards particular companies. they are renowned for being pro microsoft. so think about it. they downplay the linux penetration. it managers who are considering migrating to linux see the gartner figures. think - oh god! no way i'm moving to linux. and then they go out and buy windows 2000 which, admittidly is quite a good product. then they will probably service pack 1 their installation because they think they know better and watch as their sockets start randomly closing on them and their memory disapears into some big black hole. i suppose there is always service pack 2 - but yeah, right!

of course they are suspect!

[gnurd]

i mean how many servers are shipped running anything? usually they are turned off.
---

survey pool

[mrm677]

Another post stated that the statistics are based on a survey-- not on pre-installed sales. Thus, it all depends on who Gartner decided to survey. Did they choose only Fortune 500 companies? Did they choose universities?

Alas I may be talking out of my ass because I'm too lazy to read the details.

Re:Slashdot should not assume its readers are MORO

[mrm677]

You are a moron if you think any OS should support XML in the kernel. Obviously you don't know what a kernel is.

remember y2k

[mrm677]

I recall that the Gartner Group really hyped up the "Y2K bug".

Here is an unreliable M$ server ...

[TrollMaster3000]

http://www.icc.cc.ms.us/

Have fun guys. :)

UNIX iron gets the job done

[isotop]

I just had a look at the netcraft statistics for the top hosting locations.
And there is _not_ 1 provider on the list that only runs some MS product.
The ones that have Windows boxes usually have Linux too or some other UNIX.
Of course when it comes to intranet servers, linux will stand out that much
, i suppose MS will by far provide the most server solutions there.
I wonder what the fuss about support contracts is all about,
i never ever had one but i read that calling the MS support liens,
seldomly results in getting the job done, unless it's a very basic question
.
It's pretty much is about what you count as a sever and what not
if that means also office servers, linux should almost disappear in the statistics.

With MS dominating the market for x86 computers,
it will always be tough to a computer off the shelf that has not windows installed.

Around here...

[return 42]

Well, 100% of the machines (1) in my apartment came with no operating system at all. So in this part of the country I guess the dominant OS would be AMIBIOS 95.

Servers

[the monkfish]

the problem is that not enough companies provide linux servers so people buy the windows computers and stick Linux on them. or just get blank ones

Re:*BSD is dying

[keith.gillum]

Hmmmmm.BSD is dead/dying? Don't tell yahoo! considering thier *entire* company runs on BSD even the desktops ( well , maybe not the marketing weenies ). Hotmail uses BSD as well.