The Truth about Linux and Windows

Posted by: Steve Hamm on April 22, 2005

I’ve got a bone to pick with the never-ending stream of studies by tech research outfits comparing Linux to Windows. For starters, it seems like about half of them are paid for by one camp or another. Even when analysts aren’t on the payroll, this is really complex stuff—and useful facts are hard come by. And, beyond complexity, some studies just make me scratch my head. For example: a recent one put out by the Yankee Group. I just don’t trust its conclusions.

My beef with Yankee started April 1. That’s when I got an advance look at the press release announcing the study, and a phone interview with Laura DiDio, the analyst who wrote the report. My thought was I’d use this report to really delve down into the way such research is done and figure out how credible it is. So I not only questioned DiDio about her methods and her conclusions, but I also asked for the data upon which here conclusions were drawn. The headline on the press release is “Yankee Group Survey Reveals Majority of Enterprises rate Quality and Performance of Windows Equal to or Better than Linux.” DiDio described the survey as “totally independent, web-based, and self selecting.” She went down through her findings and conclusions.

Two things bothered me: She had surveyed enterprise and small-and-medium businesses in equal portions and presented the results combined. This struck me as odd, since enterprises use Linux heavily, and SMBs barely touch the stuff. It seems to me that the average of the two isn’t very meaningful. Second thing: The release said the respondents had deployed Windows Server 2003. That left out the opinions of people who had not installed that particular piece of software. Hmmm. Seems like that would tilt things in Microsoft’s favor.

DiDio is one of the analysts who seems to have taken sides in the holy war over Windows and Linux. She published a “White Paper” on the topic last year that was sponsored by Microsoft and posted prominently on the software giant’s Web site. Also, DiDio got caught up in the famous SCO case. When SCO first made its claims that IBM had misappropriated some of its code and handed it over to the Linux community, SCO showed samples to several analysts to prove its copyrights were being infringed. DiDio, a former journalist and not a programmer, was one of them. She reported that SCO’s claims seemed justified. She told me: “It appeared to be a direct cut and paste right down to the developers’ notes.” A couple of months ago, the judge in the case wrote that he had seen “an astonishing lack of evidence” backing up SCO’s claims. On the phone, I asked DiDio’s reaction to the judge’s statement. She said: “I can’t reconcile it. I want to see what’s presented in court.”

Back to her new survey: The press release and survey were scheduled to be released April 4. In preparation, I talked to marketing professors and other tech market researchers to ask what they thought about Web-based, self-selecting surveys. Their answer: they stink. You can’t post a survey on your Web site and let all comers decide if they want to fill it out. When I confronted DiDio with these conclusions on April 4, and told her I was skeptical of her results, she explained that the survey really wasn’t as she had described at all. It was done by a third-party survey outfit who solicited responses from a pool of qualified business computer users. “It really isn’t self-selecting,” she said.

Problem solved? Not so fast. Yankee Group sent me a set of PowerPoint slides, rather than the raw data that I had asked for. The e-mail arrived on April 5, 24 hours after the release went out. It made it difficult for me to write a timely story about the survey. (In fact, this blog is the first I have written about it.) I didn’t think much of the slide show either. One slide said “Half of Users Say Linux Deployment Is Cheaper than Windows.” You might draw the conclusion that the other half say Windows is cheaper than Linux. But you’d be wrong. The bar chart on the slide showed that 34% of the respondents have not deployed a Linux server, so have no grounds for an opinion, and only 9% said their Linux deployments were more expensive than Windows deployments.

I asked to speak to the third-party firm that had done the survey, but DiDio instead referred me to Yankee’s CEO, Brian Adamik who referred me to Phil Fersht, a vice-president of research. He called so I asked him all of my skeptical questions. He basically said there was nothing wrong with the way the survey was done and the analysis was presented. He invited me to ask detailed questions about the data, and sent some info in answer to my follow-up e-mail. One factoid: The survey had not, in fact, asked screening questions about what network operating systems people had installed. So, I guess, the press release was wrong. Also, he said, the PowerPoint slides I had questioned were unedited, and had since been improved. He also declined to show me the raw data or tell me who the third-party surveyor was.

I know being an industry analyst is a hard job. In this case, the work seems sloppy, and, to my mind, not reliable. Yankee Group once had a reputation for being a researcher for hire—distributing sponsored reports to the public. Fersht says that practice ended a year ago. Good. And good riddance. Hopefully, Yankee will work hard on improving the way it communicates with the press, too.

Reader Comments

David Mohring

April 22, 2005 8:23 PM

Laura Didio's reputation in the open source community has been damaged by her own actions.

Brent Noorda, on the statements made by Laura DiDio and others in the press about his sister's demise: QUOTE

But the journalistic integrity of any publication is defeated when articles, such as those recently appearing in the Salt Lake Tribune, include quotes from “analysts” who are completely misleading and just plain wrong about nearly every fact and interpretation. When the information provided by analysts like Rob Enderle and Laura DiDio weren’t incorrect, their statements represented speculation more fitting to a daytime soap opera than to the business section of a newspaper.

UNQUOTE

Dave Taylor

April 22, 2005 10:31 PM

Smart comments, Steve. Like any other research (think "pharmaceutical industry") it's important to know what pre-existing bias and desired outcomes there are in any IT industry analysis or projections.

To me there's a bigger issue anyway: different computer platforms have different strengths and weaknesses. I've been baffled why there's a fixation on long-term cost of ownership, when the actual analysis of COO is typically so weak or skewed that it's laughable.

phil paxton

April 23, 2005 2:07 AM

Studies like this always remind me of when the determination is made for traffic signals: it's decided where to put them, then put a meter at that intersection, verifying it, then installing it.

My philosophy has always been: gather the data, analyze it, and install the light which matches the data.

My question to people? If the data showed a traffic signal a mile down the road which appeared to have nothing to do with a particular intersection (and has only the two lanes, not a four lane intersection), would they install it there? I have yet to hear someone say, "If the data says to, then, yes!". They find things to be silly because they have preconceived notions regarding what they are looking for.

wpsmoke

April 23, 2005 2:55 AM

Didio was shown lines of code that was entitled to be in both Linux and SCO's software. It was probably public domain code or old BSD stuff. The presence identical code (and the identical comments) simply does not mean copying has occurred. You have to show that your copyright has been infringed and that SCO has failed to do.

She also says that she will wait to see what is shown in Court. This is a civil suit and you have to show your evidence to the other side before you go to Court to allow the accused to prepare their defense. A lot of observers seem to go along with this concept of "saving stuff until the trial" but the truth is it is "use it or lose it". You cannot produce anything in Court you did not make known in the alloted time beforehand.


Despite the Judges bending over backward to provide SCO with much of what they have unreasonably demanded in order to prevent there being any grounds for appeal the truth is time is fast running out for SCO and they have not shown any evidence of infringement at all never mind the blatant infringement levels that are required for the court to award the level of penalties SCO seeks.

What must really hurt SCO, Microsoft and probably Sun is that the whole exercise has just helped Linux advocates to stand together more purposefully and that IBM regards Linux as the future.

A few unbalanced "reports" will not counter the reality of the increasing acceptance of linux and the importance of open source software in general.

Tony McNamara

April 23, 2005 4:40 AM

Without the raw data no research is reliable. It is just marketing. Many research reports are worth exactly what the reader paid for them. But their purpose is not to enlighten the reader. Their purpose is to create the perceptions their paymaster wants the readers to believed.

John

April 23, 2005 6:27 AM

These type of surveys seem more like weighted advertising. I can see the headlines now.

Yankee DiDio went to town saying Linux was a Pony.

Sorry for the bad humour

Regards
John

Freddie Montana

April 23, 2005 8:42 AM

MicroSOFT is correct. Deploying windows is harder than GNU/Linux. Oops, make that GNU/Linux harder than Windows...


I am a reluctant convert to GNU/Linux. I had not the intention, time or resources to learn this new way of doing business. In addition to doing hard back-breaking work for the world's giant discount retailer, I am a young budding writer running three computers in my raggedy home office and in whatever spare time there was left, I was helping others sort out their Windows problems.

I had done so many clean install that it was automatic. After all, every known web investation was on these various boxes that it was the only corrective action availabe.

Never mind, these problems did not happen to me. I was relative safe behind a hardware and software firewall. In addition to Zone Alarm, I ran two variants of anti-virus software on my machines and above all, I never use Microsoft's Internet Explorer except to upgrade my OS, which happened to be Windows 2000 with Service Pack 4.

Everything was going fine until early one morning when I hit the wrong button and erased the folder containing my recent work from my main computer. Fortunately, for me the work was saved on the other computer: I use a little program call Folder Share that automatically duplicates designated files across several computers, but it was after this incident that I began investigating an alternative means of writing.

I automatically ruled out Apple's Mac because of the cost. I don't have that kind of money to blow on a computer and my wife would instantly send me to the graveyard after learning of my spending.

So, I turn to Linux. Years ago, I had bought a book called Unleashing Red Hat Linux, which had RH 5. I built a box out of the bits and pieces that were lying around the house and after installing the OS, I sat and looked at it, wondering what could I do now?

Fortunately the answer came from my wife, or rather her computer. Althought, her PC wasn't the newest in the house, it was the motherboard that held key. It was an Abit's KT7A with built in raid running a measily 600MHz Duron processor but there was 768MB ram and a Matrox G450 dual head video card in it.

The first thing I did was to back up all the data, from all my windows machines, to a 40GB hard drive and tucked it away for safe keeping before I began converting the computer. I stuck a 20GB and a 10GB on the IDE and used two 40GB hard drives on the Raid. Next I downloaded FC1 and by trial and error I spread the contents of the four CD's across the entire system, reserving the Raid as my home directory.

That worked fine but then I ran into a problem; I couldn't easily upgrade FC distribution unless I did a clean install. At least, I did know how to do it so I waited for FC3 before doing a clean install and in the process, upgrading the Raid to 80GB hard drives, upgrading the 20GB and 10GB HD's to 40 and 20 respectively, and adding a DVD player to the CD-Rom. Oh! I also went to the Athlon 1.4GHz microprocessor.

Now FC3 works fine. I had OpenOffice writer, and a relatively easy way to keep the computer upgraded: Red Hat's up2date works fine but above all I was back to writing with a better way of keeping my data intact. Side note: I discovered the built in dictionary, a much better calcuator and chess but back to the story.

Yes, I know I could have done the same under Win2K but I would have to specify where my saved folders would be and MS Word is known to have virus attacking capabilities. Additionally, I was thinking longterm and there will come a time that Win2K will become obselete. I am in no mood for buying a new OS: The name Longhorn should say it all and I am definetely not in the mood for buying all the trappings that will come with it: Antivirus and Anti-spyware software, etc.

But my point here is that Microsoft is correct: Deploying Linux was more expensive to me than using their product. I had no problems using Win2K. I had no problems loading the OS and I used the required tools that kept my computers safe, but once I began using GNU/Linux, I had to relearn everything. There is no consistant GUI for downloading and installing all those wonderful and free programs but everything can be down by line command and at times there is dependency issues.

Microsoft is correct: Deploying Linux is more expensive than using Windows. Not all my hardware works: I cannot get a simple thing as my radio to work, and neither would my DVD work. I cannot get independent output to the monitors as yet and ditto goes for my Lexmark printer X75 printer.

Microsoft is correct: Deploying Linux is more expensive than their products. If it was true for me with a mere three computer setup, then it must be true for a large corporation like an IBM or Novel. I can only imagine the nighmares caused by making such a transition, let alone the lost productivity, inherent downtime associated with such a switch.

Microsoft is correct: Hey wait, I have said that before BUT like everything in life, I have discovered how my computer really works. I am having fun, reading up on stuff and fixing problems on my own, albeit, there is help on the web. Microsoft IS NOT correct in this respect.

I liken this MS/Linux debate to driving on the highway: The MS camp are all in their cars whether they be a Ford, GM or Chrysler. There are speeding along passing you, speeding along some more until they meet bad weather when visibility is reduced and an accident is reported up ahead. They usually will come to a abrupt stop, waiting and wondering what to do. Think updating and a massive power outage like the one we had in the eastern USA a short while ago!

Cars are nice but they cannot go every where.

Linux is like an SUV: Big, bad and intimidating but if there is bad weather: No problem, just select four wheel drive and keep going. Snow? Off road? The same rule apply: hit the 4X4 or better yet the AWD button and keep rolling.

Linux is like the gas guzzling SUV: it takes a little more up-front to run but the ride is much better. You sit higher, get a better a better view and command a little more respect on the information super highway. If something doesn't work, open a terminal and fiddle. There is no fear of breaking the thing: the device or application wasn't working in the first place!

Porsche eventually built an SUV.

Microsoft is correct: Deploying Linux is definitely more expensive than than running with their own. Time and energy do cost and most people, for whatever reason they may have, do resist change. Insanity is repeating the same action, hoping for a different result.

So, give Redmond its claim: Now Let's move on!

Their claim is nothing but a distraction at best and an irritation at worst.

Now if only IBM would get this cell thing to market, open source it and let the boys have a go at it: Sshhh...that would be moving on.

freecode

April 23, 2005 9:11 AM

Thank you for doing the hard work of analyzing the poor research methods utilized by these studies. They actualy do more harm than good, because the studies lack controls that proper research would put into place. I began to fight against this practice when I found out that research companies will do "analysis" for money. The first sign that there is a problem is when they accept money for a study. Pure research puts controls in place to ensure accurate measurement of what you are trying to study. These studies seem to be conclusions in search of facts rather than facts seeking an answer.

Have a nice day.

freecode

kc

April 23, 2005 12:19 PM

Yankee Group once had a reputation for being a researcher for hire—distributing sponsored reports to the public. Fresht says that practice ended a year ago. Good. And good riddance.

Steve;

Do you REALLY believe that???

Thomas Frayne

April 23, 2005 1:39 PM

Hamm: Truth - summary

Didio had surveyed enterprise and small-and-medium businesses in equal portions and presented the results combined... average of the two isn’t very meaningful

She could not reconcile her "cut and paste" quote of 2003 with Judge Kimball's statement, after studying the evidence that SCOG had presented, including the sealed documents, filed a written opinion to the contrary. He stated that he was astonished that after all SCOG's publicity about the massive infringing code in Linux, SCOG could not produce a single admissible piece of evidence of infringement in Linux. (SCOG has had over 18 months to do this since it was first ordered by the court to produce this information.)

Didio did not comply with a request for the raw data, but sent a misleading Powerpoiny presentation that “Half of Users Say Linux Deployment Is Cheaper than Windows,” when, in fact only 9% said their Linux deployments were more expensive than Windows deployments.


Phil Fersht, a vice-president of research, did call, so Hamm asked him all of his skeptical questions. Fersht basically said there was nothing wrong with the way the survey was done and the analysis was presented. He invited me to ask detailed questions about the data, and sent some info in answer to my follow-up e-mail. One factoid: The survey had not, in fact, asked screening questions about what network operating systems people had installed. So, Hamm guessed, the press release was wrong. Also, Fersht said, the PowerPoint slides Hamm had questioned were unedited, and had since been improved. He also declined to show me the raw data or tell Hamm who the third-party surveyor was.

D. R.

April 23, 2005 1:51 PM

The article mentioned the "line by line copied code" which Laura DiDio allegedly saw. Ms Didio then says she can't reconcile what she saw with the judge findings (total lack of evidence brought by SCO.) Irreconcilable? Of course not.

Ms DiDio is not telling the whole truth, once again (so what else is new). The code shown by SCO under NDA to a small group of analysts, and later leaked in SCO Forum in Las Vegas, was indeed almost identical, but it wasn't owned by SCO.

The details:

Greg Lehey's rebuttal:
http://www.lemis.com/grog/SCO/code-comparison.html

Bruce Perens' rebuttal:
http://www.perens.com/SCO/SCOCopiedCode.html

Eric Raymond' rebuttal:
http://www.catb.org/~esr/writings/smoking-fizzle.html

rich wicks

April 23, 2005 2:22 PM

"Yankee Group once had a reputation for being a researcher for hire—distributing sponsored reports to the public. Fresht says that practice ended a year ago"

You believe them?

If Yankee Group is a PR outfit, how would it serve them to admit it? There job is to convince people, and in order to do that, they have to have credibility, and in order to have credbility to have to lie about their objectivity.

If you like to read, read about Edward Bernays and not that Edward Bernays was a psychologist. A good one.

Robert Weiler

April 23, 2005 3:36 PM

It is extremely gratifying that somebody in the mainstream media finally took the time to ask a couple of questions regarding information that comes off the news wire instead of simply printing it as if it were true. Having over 25 years of IT experience using both Linux and Windows, I can state that in my experience, Linux is substantially cheaper than Windows provided ones business does not rely on application software that only runs on Windows. Even in that case, businesses should belooking for alternative applications that do. The sooner you get off the Windows treadmill, the more you save, and the savings appear directly on the bottom line.

Subscriber

April 23, 2005 3:45 PM

Wonderful! This is why I subscribe to BusinessWeek! I hope you get something about this into the print version as well..

Cypress

April 23, 2005 5:15 PM

Nice article. Regarding the fact that Yankee Group is backing up everything Microsoft shoves downtheir throat it's simply disgusting. The once had a reputation. That time is over. I, personally, don't trust a word they say anymore. I am a Linux user, I used to work with Windows some years ago, so I know the REAL facts. I'll just have to let Yankee Group and Microsoft and Laura brainwash "didiot" believers...

Jozze

April 23, 2005 6:09 PM

Just hardware become more powerful to the point where we can run
on PC UNIX or UNIX like systems.
So, UNIX, *BSD, Linux become more user friendly and Windows become
more and more absolute even rapid rebuilding of Windows by putting
in more stuff from UNIX world can not stop Windows become more and
more absolute.

Han Solo

April 23, 2005 7:27 PM

The Yankee Group are worthless.

These people have ZERO credibility with professoinal IT people. Anyone who quotes anyone from the Yankee Group, and especially Didio in their articles just makes any real professional IT person roll their eyes.

Didio has so little respect I have heared many 30-50 something IT people in my fortune 40 use the nickname of "Didiot" to describe other people in common discussions.

This company and their worthless analysts are famous for how clueless they are and how much false information and wrong statements they can get published.

Corrin

April 23, 2005 8:29 PM

I think the main reason that the studies comparing linux to windows suck is that you have to grossly overgeneralise in order to get a simple answer. Any over-generalisation is going to miscategorise many people, especially when so much money is involved with getting the `right' outcome.


For example, the pro linux sites generally assume you have a highly skilled IT team that will take the extra flexability offered by linux and put it to good use. The pro windows sites generally assume you have a strictly MS trained IT team that will, naturally, struggle; and that the extra flexability means it is much easier to really stuff up.


Disclaimer, I'm much more in the linux camp, and I think Didio is attempting to deliberately trick her readers.

Donal Graeme

April 23, 2005 8:31 PM

Thank you for clearly and neutrally describing why so many people in the linux community distrust Didio and Yankee Group. I consider myself an open source advocate, but not a zealot. I fully recognize situations in which linux and other open source solutions are not appropriate. However, I become very riled when "studies" and "analysis" are presented to the public without even decent trappings of validity. I'm a data person. People I work with love formatting, graphics, frill and lace. I often wish everything was in plain text. Then I'd have information.

You, sir, have provided valuable information. Thank you.

BW Reader

April 23, 2005 9:26 PM

"DiDio is one of the analysts who seems to have taken sides in the holy war over Windows and Linux."

Maybe, but apparently so has Hamm.

David Pastern

April 23, 2005 10:51 PM

Steve,

Good report. This is what all good reporters do - they question what they are told and ASK questions.

The fact that you were given the run around by the Yankee CEO, Laura & Phil Fersht should show that their data is not to be trusted. It has been manipulated to serve a purpose. Their reluctance to provide details of the 3rd party survey conductor is worrying.

It's good to see (at least some of) the press noticing these gaps.

Dave

PS I've happily used Linux for nearly 2 years now without ANY need for a Windows based machine. Doesn't that tell you something?

Frank Daley

April 23, 2005 11:18 PM

Thank you Steve Hamm. You've done what a good investigative journalist should do, you kept digging and digging, and you've 'started' to uncover the rot.

The whole area of Microsoft's relationships with organizations such as Gartner needs to be exposed.

And if you are ever looking for a really big story, try digging into the contractual relationships between Microsoft and every major PC vendor that ensures they cannot pre-load non-Microsoft Operating Systems on PCs they sell through retail.

BDE

April 24, 2005 12:32 AM

I also take any research work of Yankee Group with a large grain of salt, and I never trust their data and conclusions. Their work seems to me very fishy to say the least.

Bax

April 24, 2005 12:36 AM

Well put. The results would have been much less skewed in Microsoft's favor if they had included all Microsoft serverside OSs, particulary Win2k which is in more use the Windows 2003 from my experience.

David D. Huff Jr.

April 24, 2005 1:09 AM

There you go taking a hard look at DiDio's work. Actually asking questions not only of her but others that are involved with her work. -Not much to work with is there?

Andrea

April 24, 2005 2:27 AM

Enlighting, i mean, how many journalists have you read questioning analysts about raw data used in their studies? IMHO, statistical studies conclusions are too easy to be skewed... =)

Aditya Narayan

April 24, 2005 2:54 AM

To make sense of the Windows vs Linux debate - one has to get down to the fundamentals.

Firstly - one should realize that any respectable TCO study should span a minimum time and at this point there are simply not enough data for a thorough analysis. Linux adoption has just started and there are not enough data points out there. Consequently these TCO studies should be taken more as predictions and be treated accordingly.

Windows was designed for an average user with no technical background. The philosophy of most of the Windows server management products is to keep everything simple for the administrator whereas the UNIX (and Linux) was designed for programmers with the goal of keeping the system simple, not keeping the life of the administrator simple. Hence many who just want to get the job done are less than satisfied with this detail oriented UNIX philosophy but on the other hand it works out very well for those who are more technically competent and like to deal with fundamentals. A natural consequence is that UNIX (or Linux) administrators cost more than their Windows counterparts. Besides - it’s not a just utilization of resources to use high end programmer level people to do day to day maintenance on a system like creating users, allocating disk quotas and the like.

And since staffing is usually the biggest component in any IT TCO study (typically staffing costs are anywhere from 3 to 4 times more than software acquisition costs over an extended period of time say 5 years) - one should not be surprised that Windows sometimes comes out cheaper. There's definitely some irony in Windows being cheaper here but the strengths of Linux lie elsewhere for the moment -


A. Linux avoids the single vendor lockin issue. (Here too there are only 2 major players - RH and Novell/Suse and very few have the skills to create their own GNU/Linux OS from scratch.)

B. Linux offers a better safety from viruses(so far - few people are writing viruses for Linux)

C. Linux is open source - one consequently has more flexibility. (Although everyone knows that very few wish to alter source code of the kernel even if they know a thing or two about that. The most anyone does is to compile their own kernels. The true flexibility comes from the finer conrol any UNIX system allows as mentioned earlier. It has little to do with the source code being in public domain)

D. Ability to customize the OS - one can remove the GUI, remove unwanted services and run a very optimized Linux server which is currently impossible in Windows. (And for all we know Microsoft may be coming up with a version of Windows that allows exactly this...)

Most other ones like the standard stability issue is a relic of the NT 4 era and currently Windows is probably equally stable on the server side.


And not all of these contribute towards a TCO directly. Even the ones that do – their contribution is not simple. Examples – single vendor lock-in. That’s a definite plus for Linux but does the TCO reflect that? Ability to customize the OS – perhaps it saves some companies from buying new hardware but perhaps for some companies – it has no other affect than the satisfaction of having set up perfectly tailored systems.

Windows has some merits and Linux has some and choosing one among them should be necessarily done on a case by case basis. Some merits can be directly compared and some can’t. Some contribute to TCO, some don’t and one should clearly realize that not everybody’s needs are the same. To put all these issues under one blanket averaged out TCO measurement (whether done scientifically or not) is to make a fundamental mistake. What I would propose is that if a TCO measure must be created for a comparison of Linux and Windows - it has to be done based on company size, internal skill set, industry vertical, company history, business strategy, and a variety of other parameters.
Then it may make sense to compare TCO of two similar companies one using Windows and one using Linux. One can perhaps analyze the TCO for historical data and come up with some valid conclusions but going forward - it does not make much sense to give a TCO estimate as much importance as it is being given in the press lately, especially given the lack of data points. Rather, a business should evaluate both Windows and Linux based on their own business needs and go with what works out best with them.

Ammar

April 24, 2005 4:43 AM

This is a great response to Laura tricks.

Thank you Steve

Morten Juhl Johansen

April 24, 2005 6:32 AM

"Fresht says that practice ended a year ago."
- Would that be around the request of the last report? I would assume they pulled in a big sack of MS money on that.
Anyway, it is remarkable that an institution with such a history and such an experienced staff simply ignores the fact that a survey published without the raw data isn't a survey, but an opinion, a postulate. It really renders the publication useless.

marytee

April 24, 2005 8:40 AM

Thanks for researching and clarifying the circumstances surrounding this survey.

Andrew Thorniley

April 24, 2005 12:54 PM

As a medical practitioner, the "research" wouldn't pass muster.
No peer review, no blinded study and so on....
Furthermore, all articles we submit have to have a disclaimer stating whether there is any conflict of interest, funding and so on, and to what extent each contributor made.

santosh kumar

April 24, 2005 1:02 PM

I want to know how i remove extended partition.
I used Fdisk method for removing windows and red hat linux fedora core 2 but it delets the primary partition and the logical partition but it cant delete the extended partition. when i restart the computer it boots up and GRUB PROMPT DISPLAYS ON THE screen.please help me how i remove it and again i want to install the windows os. and the linux on the same hard disk. Please suggest me how i will do it.

Richard Hecker

April 24, 2005 1:09 PM

The analyst claims “I can’t reconcile it." and then wonders why people refer to her as Didiot. Does she think the rest of us are as gullible as she is? These paid for studies do not make sense. They only seem to make sense when I think of them as advertising on the part of the entity picking up the tab. There are 'Truth in Advertising' regulations and these studies seem like a clever way to avoid them. The Yankee Group will need to do much better than this before they lose their reputation as "a researcher for hire."

Mark Longridge

April 24, 2005 1:18 PM

The only truth I can say is that only people who have used both Linux and Windows can meaningfully compare them.

I've used both and I prefer Linux. BTW I have used all the versions of Windows since 2.0 and I started with Unix System V in the late 80's. Tried to switch to BSD in 1995 and I didn't stick with it. Switched to Linux in early 2003 and upgraded to Fedora Core in May 2004.

For me the philosophy of open source is very appealing. Plus in the short time I've used Linux I've had uptimes as long as 180 days.

Anyways I don't immediately believe any "reports" in the media. Linux just appeals more to the programmer type and Windows appeals more to the average computer user. I don't see many people switching camps but it's just bad business to have one monopolistic company controlling everything.

I use Linux for everything and because of open source it is CHEAPER to run than windows by far. There is nothing I haven't been able do on Linux.

Mark

Howard B. Golden

April 24, 2005 3:40 PM

I hope you will keep digging on this story. (I have a feeling you will, since this entry's conclusion is surprisingly meek.)

Ian

April 24, 2005 6:53 PM

Good analysis.

Figures don't lie, but liars can figure. Not saying that the Yankee Group are liars, but, they certainly apply an editorial bias to the interpretation of their results.

Gustavo Garnica

April 25, 2005 9:25 AM

To me, the population of the world interested in computers can be divided in three groups: first those who are truly incapable of uttering an opinion on whether F/OSS software, including Linux, is better or not thatn MS software, second those who are convinced of the superiority, in every respect, of F/OSS, prominently Linux, and lastly, those who know the truth of this last assertion but who do not accept it, either because of their being on MS payroll or because they want to be in that payroll.

It seems to me that for all of us involved in software the truth is plain and simple.

John Villesvik

April 25, 2005 11:27 AM

I ran a Windows XP box with Norton Antivirus and Norton Firewalls, and still had to reload XP every two months because viruses were sneaking past my firewalls and shutting things down (I run an on-line business and am especially prone to virus attacks). In addition, my computer was falling apart - I hadn't had sound for a year and my CDRW was dodgy at best. Finally, after the last virus attack, I said, "Screw it," and loaded Linspire 4.5. It took fifteen minutes to install the operating system and reload my back-up data (as opposed to between two and six hours with XP). And all of a sudden, my sound was back (XP hadn't been able to detect the sound card for a year) and my CDRW has performed perfectly ever since (XP sometimes detected the CDRW, sometimes didn't) and my storage capacity nearly doubled. I am currently upgrading to Linspire 5.0. My only complaint is that the trackball on my mouse isn't supported (how's that for a minor complaint). If I need IE (such as for an on-line class) I use a public library conmputer. For anything else, Linspire has proven itself to be reliable, virus-free, problem-free, and a lot cheaper to run. Total cost of installation (excluding cost of my time): $0.00.

But hey, I'm just a small business owner. What do I know?

Bob Matsuoka

April 25, 2005 1:17 PM

Great article, Kudos. The tech "research" industry really needs this type of attention. Keep up the great work!

shen

April 25, 2005 1:23 PM

>the UNIX (and Linux) was designed for programmers
>with the goal of keeping the system simple, not
> keeping the life of the administrator simple

i don't agree on this. why for the purpose to keep the system simple? -- shen

fred nethel

April 25, 2005 2:53 PM

To me, the population of the world interested in computers can be divided in three groups:them who can count, and them who can't.

Charles

April 25, 2005 3:22 PM

It is amusing to me the lack of core technical knowledge by supposed IT personnel in one or two of the posts above.

"Linux offers a better safety from viruses(so far - few people are writing viruses for Linux)"

Wrong. Linux offers better safety from viruses because the applications do not have ready access to the kernel of the operating system. In other words, Linux is designed NOT to let the inmates run the asylum. Microsoft allows many key applications with outward exposure this privilege, something that is taken well advantage of by malware authors.

" The philosophy of most of the Windows server management products is to keep everything simple for the administrator"

Until something goes wrong, which is inevitable with computers of any sort, or any OS. Then, skill and ability to get help come into play.

Have you ever searched Microsoft's Knowledge Base? You may as well pull out a Ouija Board.

Linux preferred mode of support? It's called Google. And that is before you ever try to contact a vendor. Most problems are solved with a simple web search.

Mike Pfaiffer

April 25, 2005 3:28 PM

Just so you know at the beginning I have been a Linux convert since the mid 90's. In the past I have held jobs as tech support for a LARGE ISP, a programmer (more than a few times), systems analyst, many other related professional positions. Currently I'm taking a break (sort of) and writing articles for a non-Microsoft oriented computer magazine. Like quite a few of the others who have commented thus far, I started my career path in the days of CP/M and DOS. I have used many different computers ranging from the microcomputer all the way up to a mainframe. I have used Windows as well. In this day and age who has not used Windows?

Having stated my bias up front the readers can give the appropriate weight to my comments.

Sir, you have done what few others in the media have done. You have actually asked for more information and proof. If a piece is an opinion piece or a piece relating an experience of the author then it's usually fine to ask "light weight" questions. However in this case the author wrote about the analysis of some data. Surely it is reasonable to ask to see the data and how she came to her conclusions. I hope you continue to ask the hard questions...

FYI, in my experience the cost of running Linux for anybody with minimal computer experience can be as little as the cost of time and hardware (which they would have to invest in order to run Windows anyhow). As an aside, a non-technical friend set up an 80 year old great-grandmother with a Linux box. It was a gift from her grandkids. He set it up so it is easy for her to use. Basically "click here to write a message to your grandkids". After four years she has never had a problem with the machine. She did pay him $10.00 to come in and install flash when she discovered how to use Mozilla, but that was it. Unlike Ms. DiDio I'm willing to introduce you to my friend as well as others who have been using Linux happily for quite a while.

joeldg

April 25, 2005 3:49 PM

As a prior windows developer who has switched to linux as of several years ago I can tell you that MS windows is by far more expensive for me..
I recently got a laptop and decided to dual-boot it and play around with some windows apps.. It was not long before the whole thing was infected and it took me hours and hours to clean the machine of all this garbage (time which I could have spent being productive, or at least enjoying myself playing a video game)..
This is a scenario that would never repeat itself on a linux platform.
Your browser cannot take over your computer and put you at risk.. A regular user could never infect and destroy an entire computer thanks to permissions and just the entire 'mindset' of linux.
With windows, you not only get to pay for the OS, you get to pay for anti-virus solutions as well as paying for any software installed and in the end your business is still at great risk thanks to ubiquitous spyware.. (think of the halflife2 source-code debacle) and any new viruses that might come and randomly destroy your work.
It is nice to see a mainstream reporter finally standing back and taking a look at this blatant PR and helping to stomp these paid-reporting firms in the face..
Perhaps there is a need for a real un-biased reporting firm that is not under the thumb of a corporation (on either side).

Jay

April 25, 2005 3:58 PM

/rant on.

Combine "con" and "insult" and you get consult. Remember that.

Unless you've used linux in an enterprise day to day in my opinion you are not qualified to judge it's merit. Laura Didio's article is like reading 1/2 of Freud's works, talking to 2 people he counseled part time and calling myself a shrink. Just because i'm familiar with the material and 2 people told me i'm correct doesn't mean I know what the hell i'm talking about. People use what they are comfortable with. I prefer linux for a domino environment, but prefer my AD for my users' domains. Fortunately the people above me in my dept. aren't swayed due to fly-by-night reports from "analysts" who are poorly completing the task of reading interpretation.

George

April 25, 2005 4:25 PM

What everyone here seems to miss and what Microsoft is so absolutely aware of is that while it is true that these people and their bogus studies have no credibility in the IT world, so what?

The intended audience is not IT pros who know better but their pointy-haired managers who hold the purse-strings and make the decisions as to what to buy.

DiegoM

April 25, 2005 4:29 PM

Good Article. There is no data to support the claims fo Yankee Group, so it's as useful as me saying "I know pigs can fly, I've seen it".

I am neither an M$ or Linux zealot. As someone said in a previous post, each one has his strengths, and I like when I depliy a strong and secure infraestructure showing both MS Windows and Linux servers. We have to stop thinking this as a holy war. You can have the best of both worlds, coexisting peacefully.

jusdisgi

April 25, 2005 6:20 PM

This is in response to Freddy Montana (the guy with all the "Microsoft is correct:"'s in his post).
They aren't correct, or at least not for the reasons you cite. You say they were right, that it cost more to use Linux than to use Windows. And that if this was true for you with 3 computers at home, than it is likely true for a company with many more computers.
There are (at least) 2 reasons this logic is flawed. First, you meant to say, "It's more expensive to switch to Linux from Windows than to keep using Windows." This is worlds apart from one being cheaper to use than the other. In the latter case, you assume the startup costs and licensing and (most importantly) learning costs equally. The way you are evaluating, you consider all those things free for Windows, and costly for Linux. That's just not a fair comparison.
More importantly, though, the evaluation of cost at your home, where you have an IT staff of one person you can't fire (you), who apparently already has significant Windows experience. That is to say, you have already made a massive investment in Windows training. This is completely different than the cost evaluation of a company, where they can hire whomever they want. Perhaps this is equivalent to a company with an existing IT staff consisting entirely of MCSE's, and management that isn't willing to replace them....but I'd say that business is making a big mistake. Here's a hint, folks; if you want your Linux rollout to go smoothly......hire people who know Linux. If you have an MCSE-based staff, and you're switching to Linux....fire them, and hire Linux geeks. You wouldn't take your new Volvo to the Toyota dealership where you took your last car, would you?
Linux admins are somewhat more expensive per hour than Windows admins. The amount of work, in terms of how much server output they can maintain on a per-admin basis, is vastly higher. Thus, for a corporation, none of your logic adds up at all.

Gee. I sure wish I could figure out how to make paragraphs on this blog software...I tried typing in just plain text, or with html tags, and either way it just gives me a single block of text in the preview. Oh well. Imagine the paragraph breaks.

jdog

April 25, 2005 8:12 PM

Freddy Montana's post just gave me what an independent study group SHOULD do: compare the costs of switching from Windows to Linux as opposed to Linux to Windows.
If this (hypothetical) study ever does get published, I'm thinking the conclusions will say that it would be about equal.

Hasse

April 25, 2005 8:50 PM

Great article! It is very importat that this reach the avrage computer user since people have a tendency to belive everything that is printed in the newspaper.

It's 3 am so I don't have the energy to write a long post about why Linux is more cost effective, many people have mentioned reasons earlier in this thread. But I will say this. Linux is in many ways a superior OS and if deployed correctly it can (and will) save your company money and give you far better reliability.

One thing that hasn't been mentioned here is that open source products can be supported by many companies. This means that the open source vendors has to compete with eachother in terms of providing the best support. THIS is a great advantage for Linux.

(sorry for my bad english)

Alan Whitmarsh

April 25, 2005 10:03 PM

Full marks to the Yankee Group for covering this issue. The realities and flaws of Linux need to be communicated to the public, and they have taken a stand to do this. DiDio gets far too much criticism for voicing her opinions, and Mr Hamm has failed to highlight any flaws in this survey methodology. Thanks Yankee - you have a great reputation for speaking the mind of the REAL user and have the guts to come out with these findings - please keep it coming, I have read your research for years, and always find it unbiased, credible and takes a stand there others (e.g. Gartner, IDC etc) choose to sit on the fence. As a CIO for many years, I can vouch that these results are spot on.

Leon Brooks

April 25, 2005 11:27 PM

Good to see mainstream press asking real questions and taking public note of the lack of real answers.

Keep right on calling these con-artists' bluffs and we'll keep right on reading you. (-:

Vivek

April 26, 2005 2:10 AM

Well done!

Impressive analysis of report

I used Linux (and FreeBSD) over cost saving (1) and quality of software (2) issue as long as these features exists I will be continue to use Linux :D

Richard

April 26, 2005 3:44 AM

To quote a previous comment: "A natural consequence is that UNIX (or Linux) administrators cost more than their Windows counterparts. Besides - it’s not a just utilization of resources to use high end programmer level people to do day to day maintenance on a system like creating users, allocating disk quotas and the like."

.

Perhaps, but the difference comes in UNIX's flexibility. Is it worth hiring a number of cheap Windows admins to sit through manually entering the 1000s of new student users into the system, or do you get the competent UNIX admin to use a little script and have it done automatically and reliably from the new starters list?

.

Microsoft are realising this, so later Windows versions allow things like scripting - but how many people nowadays automate things like their new user procedure? How many people have Windows Scripting enabled given the virus risk? A good UNIX admin can have so many of these things scripted and that it enables a larger number of machines to be managed by fewer people.

.

Things are changing. The Linux camp are realising that they need to attract the smaller time administrators so offer easy to use admin tools for these tasks. Microsoft now offer scripting and are hardening their system to improve security. That hardening is interesting, as it's making the system more UNIX like and therefore less "User Friendly" (for definitions of User Friendly like "familliar to old school Windows admins"). Windows admins are suddenly having to learn multi user concepts and security!

.

With Windows and Linux both hurtling towards user friendly and powerful perhaps the advantage will eventually come down entirely to purchase and upgrade price.

Joe

April 26, 2005 4:48 AM

Great piece of journalism! Great work! It's refreshing to see a reporter actually spend the time to investigate a story rather then just regurgitate what they have been given. I hope more of your collegues follow your lead.

Another Steve

April 26, 2005 5:22 AM

I got a bone to pick too, Steve. It's with bigtime media companies creating "Blogs" so they can seem like they know "Tech". Since Blogs are now more hip than magazines, not only will it appear as though Businessweek knows technology, but since its a blog...they no longer need to worry about facts or proper research. Let's just post Steve Hamm's person opinion of Laura Didio. Brilliant! Nice work Businessweek!

I also don't believe that you spoke with marketing professors or other researchers. If you did, please cite your sources. Also, how did you screen your "research experts"? Please tell me that you at least culled your research contacts using a reputible clustering method, such as K-means. God, don't tell me you settled on something as arcane as factor analysis to build your "experts" clusters.

Listen, show me any quant research and I'll point out flaws in the process to obtain it. That's the way this business is.

I suggest instead of ripping on Laura Didio (or her research), that you spend your own money and try to do the research yourself. You'll find it extremely difficult (and expensive) to screen participants in this sort of research. But you'll have fun learning all about statistics.

Geoff Stevenson

April 26, 2005 5:55 AM

Excellent article. And some excellent points made by various posters. For a few beautific moments I imagined a world in which it had become the trend for interested parties to demand integrity from organisations, like Yankee Group, who sell themselves as impartial reporters of facts. The Yankee Group might execute research, but as long as they won't release that research as raw data, their principal business function is quite obviously the writing of Advertising copy. Any organisation setting itself up to do what Yankee Group claim to do - disseminate impartial facts - would attach considerable import to the attributes of credibility and the professional respect of their peers. Yankee Group are evidently making so much money from advertising by stealth that credibility and respect have long since fallen by the wayside.

Aad-Jan

April 26, 2005 7:22 AM

Thanks George. You seem to be the only one who get's it. Just to make sure I'll quote you. AND add some.

"What everyone here seems to miss and what Microsoft is so absolutely aware of is that while it is true that these people and their bogus studies have no credibility in the IT world, so what?"

"The intended audience is not IT pros who know better but their pointy-haired managers who hold the purse-strings and make the decisions as to what to buy."

What's more. Those managers do NOT read this blog.

It's sad but ... 'No manager ever got fired for choosing M$'. And that attitude is still rife in managers world.

Steve, thank for the article and the blog.

Jojo

April 27, 2005 8:38 AM

I had tried to install a Linux OS, but its much harder to install compared to Windows. In Windows 2K or Windows Server 2003, I can install windows after setting my Server to boot from CD and let it do the installation unattended. Not sure if Linux can do this.

For the debate on the cost of Deployment, Linux will be cheaper if you just started to deploy a new network infrastructure, but if your going to migrate or switch to Linux, then it will be a differrent story...

About security? Well, I cannot compare but what I can say is my Windows boxen is secure... :)

Gregg Dippold

April 27, 2005 2:24 PM

When Laura was at Giga now Forrester, I had the misfortune of sitting through one of her pompous presentations on Windows. She is a journalist pretending to be an analyst. She dodged every technical question after we endured her sonorous wheezings. She is living proof that confidence and political acumen are the primary skills for analysts.

Pradeep Kumar

April 27, 2005 3:23 PM

Steve Hamm embodies the old saying "Never let the facts get in the way of a good story" .... except this is a bad story Steve. His bias is transparent and he should title the article "My opinion". Steve you get an F for full disclosure.

Research firms are not non-profit organizations, therefore they charge for published research Steve. Perhaps you need a crash course in business 101? Stop expecting free handouts and entitlements, Steve -- your employer orbids it anyway.

He seems to have an ax to grind with this particular analyst firm and analyst. A serious journalist would not devote an entire lead article to character assasination of a person and their employer. He rarely mentions the business and technical issues of deploying Linux, only to criticise the authors of research.

carlos

April 27, 2005 6:47 PM

Response to AnotherSteve:
not a dumb comment given who you appear to work for (webtrends) where a prominently listed "Business Solutions Partner" is... you guessed it, Microsoft.
lol, not too difficult to see who butters your bread.
though in fairness, stlong, might not be who he claims to be, and may not be associated with webtrends.

James Runn

April 27, 2005 9:59 PM

About viruses - one should be clear - there is technically nothing that precludes a virus being written for Linux. So please don't fall for what some people want you to believe.

The first post is right - Linux is only safer for now because not many are writing viruses for Linux.

Over time - nobody knows - maybe Linux will get equally hit.

shhf

May 2, 2005 7:26 PM

James Runn says : "The first post is right - Linux is only safer for now because not many are writing viruses for Linux."

I don't think you understand security at all.

Have you even used Linux? Do you know the security features found in Linux which are completely lacking in Windows?

I use Windows, Linux, and BSD. The major thing Windows offers is "ease of use". The problem is, MS and its unusual way of handling security issues.

(1) It ALWAYS takes MS a minimum of 1 month for a security patch to come out, once they are informed of an issue. This is a fact.

(2) MS will judge if the issue is a security threat or not. ie: If there is no known exploit, then they won't bother. (so an issue is just left wide-open).

(3) MS treats security as a public relations issue. ie : They only do it to satisfy customer's "whining". More often, these are just slap-stick band-aid solutions. eg : Buffer overflow issue...They just slapped on Non-Exe bit. Another is anti-spyware.

If they were serious about security, they would've rewrote Windows from scratch. Learn what is wrong with the code, and what can be done better...But no. They're just put out band-aids.

So how's this different from open-source solutions?

Quite a bit. 1 month is plenty of time to create/deploy an exploit and by the time MS issues a patch.

Some open-source projects are very serious about security. The best example is OpenBSD. They deliver a patch under 8 hours.

Think about this...If you have a security issue and you are connected on the web, do you really want to wait 1 month or 8 hours to resolve a security problem?

So why isn't there viri for Linux?

Because it works in a completely different way to Windows. Unix and Linux users need to be aware of rootkits. But if you properly setup your system, the probability is extremely low. (far lower than it is on Windows)

The security features in Linux and BSD make it harder for blackhats to write something nasty. Its easier just to go after Windows.

The most common cause of system compromise is actually the user mis-configuring the system.


Jojo says : "About security? Well, I cannot compare but what I can say is my Windows boxen is secure... :)"

Pfft...Give me 20min, and I'll show you how "secure" your boxen really is.

How is Linux hard to install?
I'm up and ready to surf the web in 20min.

I've seen 60 year olds use Linux without an issue.

I really do question your ability to secure any OS if you can't even handle ones outside of Windows.

Sili Valley Guy

May 4, 2005 11:57 AM

On a variation of Steve Hamm's comment about Yankee's DiDio and her shoddy research methods, is tracking how accurate the analysts are in their predictions.

Below is a prediction(1) from Gartner Research Note G00124887 “Our Top Predictions for 2005, and Beyond” which is part of the annual Predictions series that got it big time wrong on telecomm acquisitions. A number of interesting points about this prediction:

• Qwest was never even mentioned
• Verizon was dismissed as a potential buyer of AT&T or MCI
• BellSouth never made a play for AT&T or MCI
• Is was SBC that bought AT&T, not BellSouth
• This prediction was wrong within only a couple of months

What does the analyst being so wrong mean for clients relying on the recommendations of the analysts?

(1) I am not violating Gartner's copyright by including this extract from a Research Note. When the Research Note was published it was freely availabile to non-clients as part of Gartner's marketing efforts. The larger firms (e.g., IDC, Forrester, Gartner) usually put out a series of "Predictions" papers every year in November or December and post them free of charge.

Prediction: In 2005, AT&T will be acquired by BellSouth and MCI will be acquired by SBC.

A number of forces are coming together to make this prediction possible, and then improve the likelihood that it will occur.

Regional Bell operating companies (RBOCs) such as BellSouth, Verizon and SBC no longer have regulatory restrictions concerning what network services they can offer. In the past, they were unable to acquire interexchange carriers, such as AT&T, MCI and Sprint. The RBOCs wish to be leading fullservice providers.

RBOCs have the financial strength to conduct mergers of this size. AT&T is more likely to be acquired by BellSouth due to past relationships and negotiations. The direction and financial situation of Verizon — and the apparent lack of interest of its key executives — make it less likely that Verizon (compared to Bell South or SBC) will make a major acquisition. Therefore, MCI is more likely to be acquired by SBC. AT&T and MCI are No. 1 and No. 2 in market share in long-distance services to the Fortune 1000, respectively, but their long-distance revenue has been declining more than 10 percent per year, primarily due to RBOC inroads in consumer long-distance voice. AT&T and MCI are financially constrained to grow or acquire new businesses to offset the revenue declines. Although the RBOCs would rather wait for such mergers, other companies' interest in MCI and AT&T will compel them to move quickly rather than later.


______________________________

One of the areas that a lot of folks grumble about is that nobody calls Gartner out when it is wrong. Perhaps something that BusinessWeek can do is periodically review Gartner’s predictions and rate them on accuracy.

______________________________


Sorry for being anonymous, but I work for a technology vendor and I have legitimate concern that Gartner will retaliate against my employer because of this comment.

James Runn

May 10, 2005 1:32 AM

There's hardly any value in these posts by bigots Some of the balanced ones are good since they present a case and let you judge - but for some of the others..the lesser said the better.

Think of those posts that have only positives about Linux and only negatives about MS: such a stand is a logical impossibility.

And about viruses I must say again that Linux is not immune from viruses and many other kinds of security vulnerabities and whoever is challanging it ought to consult his nearest security expert to understand why.

Daniel Lott

May 15, 2005 2:47 PM

It seems that Yankee Group is STILL a distributor of sponsored reports (although they're now desperately trying to hide it)

WW

July 8, 2005 4:49 PM

OK we all know that Linux is more robust than Windows but what about the distros, why does Red Hat have to make "its" Linux different from the others? We all know the catch is: making a O.S. dependent on the vendor. That´s when Linux turns to be as bad as Windows.

8thstreetlatinas

August 13, 2005 9:23 AM

This is in response to Freddy Montana (the guy with all the "Microsoft is correct:"'s in his post).
They aren't correct, or at least not for the reasons you cite. You say they were right, that it cost more to use Linux than to use Windows. And that if this was true for you with 3 computers at home, than it is likely true for a company with many more computers.
There are (at least) 2 reasons this logic is flawed. First, you meant to say, "It's more expensive to switch to Linux from Windows than to keep using Windows." This is worlds apart from one being cheaper to use than the other. In the latter case, you assume the startup costs and licensing and (most importantly) learning costs equally. The way you are evaluating, you consider all those things free for Windows, and costly for Linux. That's just not a fair comparison.
More importantly, though, the evaluation of cost at your home, where you have an IT staff of one person you can't fire (you), who apparently already has significant Windows experience. That is to say, you have already made a massive investment in Windows training. This is completely different than the cost evaluation of a company, where they can hire whomever they want. Perhaps this is equivalent to a company with an existing IT staff consisting entirely of MCSE's, and management that isn't willing to replace them....but I'd say that business is making a big mistake. Here's a hint, folks; if you want your Linux rollout to go smoothly......hire people who know Linux. If you have an MCSE-based staff, and you're switching to Linux....fire them, and hire Linux geeks. You wouldn't take your new Volvo to the Toyota dealership where you took your last car, would you?
Linux admins are somewhat more expensive per hour than Windows admins. The amount of work, in terms of how much server output they can maintain on a per-admin basis, is vastly higher. Thus, for a corporation, none of your logic adds up at all.

sundaresh

December 28, 2005 11:58 PM

Linux and the bsd's are a boon, to the programming
community since they are really free.Microsoft is
a business.Quality vice, gnu,and opensource s/w are to my knowledge, exceptionally good. The core tools like gcc etc are really way better than microsoft compilers.

Imago

January 2, 2006 12:21 PM

the "Truth about Linux and Windows" is over here :)

Scott Dennis

April 6, 2006 10:11 PM

WOW! This is a joke for those who think it is easier and cheaper to install MS products vs. Linux. Case in point. I have installed MS for years and even worked level 3 tech support for MS. I used to think that they ruled however I was hired to setup two DELL 2650's using Microsoft’s Small Business Premium 2003. Having never installed SBS I went to town thinking it would be quicker and easier than Linux FC3 and would allow the client the ease of use. After two weeks of research, trial and error and countless installs of the OS I was finally able to get the server up and working however now I needed to configure it. Still working on this one. The setup needed to have all the regular stuff like good security, groups, policies, users, fire wall, anti virus etc... Oh and needed to have email and web hosting. The information that SBS was asking for was weird in the wording and one had to pay very close attention to what you were doing. Time taken to achieve this without any experience of using SBS was well over 200 hours so count the cost on that one when our client is getting charged $85.00 per hour total cost to the client $17,000.00 not to mention the costs of the servers to handle it. Each server cost almost $4,000 each but let us be fair here this is a premium setup so we should at least not consider the hardware costs. Hardware setup included 4 GB ram, raid 5 raid controller, 2 processors, three 80 GB hotswap harddrives, dual hotswap power supplies, DVD, Dual NIC’s and floppy.

Now let’s look at Linux Fedora Core 3. I installed this on a Compaq Proliant 5500 with four 550 MHz processors, 3 GB ram 6 18.2 GB hotswap harddrives, CD ROM, 40 GB tap backup, floppy, dual hotswap power supplies, raid 5 raid controller, DVD ROM and dual NIC’s total cost for this server $678.00. I had never at this point ever installed Linux or a server software before except for Windows NT 4.0 in school which the instructor did most of the work so I thought as this would be a machine I could use to learn on I decided that I would try the Linux thing as it was so well touted for how great and stable it is. I was not sure so I wanted to know for myself. I was very pleasantly surprised when after configuring the server and putting in the CD how easy it was, everything installed with minimal help and information from me. I felt like I was installing MS Windows it was so easy. Now to look at the setup of the server as I wanted to host my own web server and email as well as provide my family with a secure way of storing data. We have a network which includes my machine, two laptops, my wife’s machine, 3 machines for my children a test machine, a production machine and the server. I was totally amazed at how easy it was to find help and the documentation for setting this all up was very well written cant say that for MS have to get through all the funky wording used which can be very confusing. I was able to get the server up and running with a fire wall, anti-virus, office suit and a lot more junk I didn’t need kinda like MS, but then I was used to getting more than I really wanted so no big deal. Total time to get the server up and running was less than 70 hours actually it took me only 3 days and in that time I also installed SCO’s free OS for a desktop machine. Total cost to me in lost wages and production at $85.00 per hour was only $5,950.00 only 35% of the cost of the MS setup. Now that is savings.

Oh now lets look the cost of downtime due to viruses, having worked for MS level 3 tech support on the MSN project I am very familiar with viruses and the removal of them as I was one of 3 people instrumental in resolving the Blaster virus issue. I was very tired of getting viruses on my MS machines this is what caused me to explore other avenues Norton and MacAfee were just not cutting it, all the updates and still getting a virus on my production machines ever 6 to 9 months was just to much of a headache to continue using MS. Now before you go on to say that MS is used by so many and that everyone hates MS let us just keep this in perspective, I for one do not hate MS and it is because of the way the OS’s kernel is written why MS has so many attacks. So I decided to give Linux a chance and I am very, very happy with my Linux Fedora Core 3 server, I seem to never have to do anything but the security updates which I have to do on MS and is much quicker and just as easy all you do is click on a little red icon just like MS’s little blue windows update icon. Yes it is different and I do not use the command line so I would have to say the GUI is just as easy as MS once you know were stuff is and as for me it is easy to find things albeit the names are funky but it is no different than it was when I learned how to use MS Windows 3.1. It is just another learning curve like learning any new software. I do have this to say for Linux it is easier to change things than it is with Windows SBS 2003 and if something doesn’t work or breaks it doesn’t mess up my whole install. I like how modular Linux is. It simply works and is very stable I am now a true convert to Linux and love all the free goodies however I must say I still have to have my MS machine for the games I have come to love I only hope more people will program them for Linux soon.

I will still use MS for my clients when they insist but I am steering them all away from MS as it really is more cost effective for them and I decided I will eat the cost of retraining them as it is in their best interest in the long run. Oh and one more thing I just found out that Linux even has a free ERP program and I can’t wait to install it as it will greatly increase my business.

Oh and Linux is by far less expensive than MS to setup install and maintain and I for one don't care what anyone’s agenda is. I care about what is best for my clients and what will save them the most in the long run even if it means I must loose some money helping them learn new software as it makes my job easier and I can make more money in the long run.


Good work Guys and Thank you so much.

Scott Dennis
scott@i-tadd.com

Barona1 Barona1

June 9, 2006 2:32 AM

As a Computer Engineer, Linux is faster, easier and stronger. Do not be fooled by journalist thinking they are pros in the content of their report.

Robert

August 17, 2007 3:26 AM

http://www.sabayonlinux.org/
and knoppix linux booting from dvd or cd drive ones

Not sure about business I am just a student college kid etc..
I use it for games and sometimes school work, plus windows dual-boot for the non-linux games, that can't yet run under linux, I yes I tryed vista but prefer linux over it , and prefer windows xp to that of vista, vista is not really a good system for home gamers, people etc..

Kevin

July 7, 2008 6:36 PM

I would love to have a discussion about the merits of Linux and Windows, but these days you could not convince someone that 2+2=4 let alone that an Operating System is a tool and not a way of life.

Jason Charles

August 22, 2008 10:05 AM

Well,time is a great leveller.
Reading this 3+ years on from publication date, all of my company's client firms now run 100% Linux + Unix, all converts from Microsoft.
Thats 19 SMEs and around 900 devices in total.
All but one dont use external support now and get along just fine. Staff costs seem to be same, I see no more/less IT staff on my travels. Stress is certainly lower, a surprise to me. Initial cries of "help" "this is crazy" stopped after the 1st month, then all the competitors went the same way.
PS-I sell HVAC for a living. I went Linux at home 6 month ago, no going back now :) I feel free

SA

February 18, 2009 1:28 PM

I have been a systems/network admin for over 11 years. Over this time, I have set up Windows and Linux servers in parallel, and Linux wins hands down. Why? Because it is stable, inexpensive, and easier to maintain in the long run. I have started writing an article mainly out of pure frustration of using Microsoft products, and finally decided to convert over to Linux at home as well as my business. I am sorry that I took so long to finally make the move from Windows to Linux.

http://members.apex-internet.com/sa/windowslinux

Pawan Kumar

February 22, 2009 3:01 AM

I have Red Hat linux 5 customized with e-court, I have been directed to install Window XP, but there is no any harddisk partition instead of C drive, tell me procedures how can I do it.

Lovezune

May 12, 2009 12:53 AM

I was not sure so I wanted to know for myself. I was very pleasantly surprised when after configuring the server and putting in the CD how easy it was, http://www.zuneconverter.org everything installed with minimal help and information from me. I felt like I was installing MS Windows it was so easy.

Post a comment

 

About

Bloomberg Businessweek writers Peter Burrows, Cliff Edwards, Olga Kharif, Aaron Ricadela, and Douglas MacMillan, dig behind the headlines to analyze what’s really happening throughout the world of technology. Tech Beat covers everything from tech bellwethers like Apple, Google, and Intel and emerging new leaders such as Facebook to new technologies, trends, and controversies.

Categories

 

BW Mall - Sponsored Links

Buy a link now!