The report said the operating system and PC represent less than a third of the total cost of ownership and that migration should only be considered in a few situations. Michael Silver, VP and research director at Gartner, said in a statement that other costs such as labour, training and external services should be taken into account.
He said organisations should compare the costs and savings of a move to Linux with the cost and savings to upgrade to a newer version of Windows as the total cost of ownership will vary depending on which version of Windows is being considered, according to Gartner.
"Enterprises running Windows 95 will likely see more benefits by a move to Linux than will enterprises using Windows 2000 or Windows XP. Windows 2000 and Windows XP include more modern technology than Window 95 and are generally more stable and incur lower costs," said Silver.
Gartner revealed that while Linux has had success in the server market reducing costs, the same savings cannot be achieved on the desktop.
David Smith, VP and Gartner fellow, said in a statement: "Many servers are dedicated to running a single application; in many cases, it has been relatively easy for enterprises to replace specific servers, such as a web server, and implement Linux."
But the environment for Linux on the desktop is significantly different, he said.
"For those users, migration costs will be very high because all Windows applications must be replaced or rewritten," said Smith.
The report comes in the same week that a Microsoft-sponsored survey of just 12 companies claimed firms could save up to 28 per cent by developing certain programs with Windows rather than Linux.
Talkback
When bored do inane research.
Upgrading to a Ferarri from a Ford is only less expensive if you do not buy a Mercedes on the way....
It seems the central issue here was the replacement of Windows applications.
The interoperability of OpenOffice applications has been totally overlooked...
I wonder who really paid for this study?
you guys are n008s, linux is 100% free and open source. yes, you may need to learn how to use linux, but it really isn't that hard...at all, i picked up the red hat linux 9 bible and finished it in about 4 days. It has come quicker to me because i have been using linux for years, but none the less it is still cheaper than windows. Did you even think about all the bugs and all the times windows crashes? it is completly unreliable. the bottom line is linux is free and will always be better than windows!!!!!!!
Sorry, guys, they are 100% right. Windows is cheaper than linux for the DESKTOP. The cost of hardware and OS are nothing compared to the cost of training your people on a new OS. 4 days per employee to learn the "new system" in a company of 1000 users is a MAJOR cost consideration. Linux still has a little ways to go in order to be a cost effective replacement desktop for companies. It needs to be a little more intuitive, have a few more configuration utilities, and it needs to have some renaming done (rename kppp to "dial up service"). Of course, Linux still beats Microsoft with a pretty big stick when it comes server deployment and security performance,
One would choose instead to agree to the fact that judging costs is extremely difficult and can only really be viewed with respect to timing and specific cases. I.e., some interests deploy a limited amount of applications, many of which are internally created - they view an Operating System (OS) as a mere platform. Others are multi closed source application deployment zones - they see an application as an unbreakable link to the OS.
Judgements are made even more difficult when current direct costs intermingle with current indirect costs both of which have to be adjudged against all of the future benefits that can be derived from particular choices. Its not just about current expenditure/savings, but about other issues; potentials, that are future based, such as increased productivity, increased creativity, increased adaptability, increased control, increased trust, increased interactivity, and others, many of which are difficult to judge from mere and unconnected/isolated business formulas.
For so many organisations, the type that do not only pay for software, but must deal with administration, security, development and other such issues, the greatest fear that relates to migration would usually surround the direct one time impact that their chosen migration path has or would have on current budgets and the risks associated with such impacts and the effects of any migration on current productivity and therefore competive abilities. In short, current competition and fixed budgets, tie them to current expenditure patterns and traditional methods irrespective of changes in environmental conditions.
In the long term however, many would agree that the advantages of moving to the open platform "cost + benefit wise" would be signifcatly more than running most closed source alternatives.
So the choice for them, is not whether one product is better in absolute terms than the other, but of if they are better off leaving their capital taps slowly dripping money over the burning flames of companies like Microsoft over the very very long term, or throwing a cup of water to seal off such losses in a one time effort.
Yet, though one believes the above, one also beleives that everything is a matter of stages and time. Witnessing a period -a mere 2yrs ago, not of kppp, but one when setting up pppd for a novice was a day or two's effort, one does suspect that with a greater proportion of the world in the development stage of their economic cycles, the adoption and therefore adaptation of Open Source products would make most of such deployments cost advantageous.
Such advantages may be as quick in coming as the development of KDE. The question will be, should organisations allow themselves not to be prepared for them when they do come...
The report is biased and obviously done for Microsoft. What is to argue.
Just to argue:
1. The Micro$oft paradigm will not work.
EG Sell a closed system, with closed tools and expect applications made by outside developers to work. If it doesn't work, the developer will make it limp enough to sell, then blame the problems on Micro$oft. Microsoft will of course blame it on the developer.
If all TCO issues were included I'm sure that Micro$oft is the more expensive.
2. When a study is done using Micro$oft are all costs truly included.
EG The cost for each desktop, each server, the cost for connecting each desktop to each server.
You know the nickel and dime stuff Micro$oft
does. Pay $xxx for a Windows server license, pay $xxx to make it a print server, pay $xxx as an additional amount to connect each desktop to each print server. Pay $xxx for upgrades for same after Micro$soft decides to change version after a year of so.
Same thing for Webservers and multiproccessor servers.
3. Lets talk about Windows support. Of course basic support was dumped by Microsoft with version 95. If you remember they told the OEM's it was their problem. Even stopped including doco's.
No basic support. Give Microsoft a call sometime and see how much support they are willing to give you. Other support ranges in cost from $100's to $1000's of dollars per hour. Are these costs included in TCO studies.
4. How about the cost$ that results because Micro$ofts operating systems are full of holes. Do your really think that these costs are included in TCO studies comparing Micro$oft and other OS's.
As stated above the paradigm doesn't work. Someday soon there are going to be several millions(millions) of software engineers in the world. Most of them not working for Microsoft.
They will want to use software they can afford, this is not Micro$oft. They will want to use software they can repair, this is not Micro$oft. They will want secure software, this is not Micro$oft. The will want reliable software this is not Micro$oft.
Re "Upgrading to a ferrari from a ford..." etc...
Most people wouldn't want to drive a ferrari all the time.
I'm as sceptical as the next anti-M$ person, but the fact remains that linux on the desktop PC isn't in the market. Where I see it potentially making headroads is through the network admin and applications markets. Most corporate users use maybe 5 apps on their systems. If you remove the need for these apps to be MS-based, users don't care which desktop they're running. But IT managers still do, so you need to beef up the enterprise desktop admin tools too.
Then, who cares what OS you're running? That's how it always should have been. To use your car analogy, who cares what engine management system they have in their car? Most people don't even know.
Interview with someone who saved money with Linux:
Reporter: How did you save money with Linux?
Average User: Well I'm about as cheap as they come. So I went down to the used PC shop and bought a 2 year old Dell for about $200 got a Linux distro mail order for about $5.
Reporter: Was it hard to put Linux on a PC?
Average User: Not at all. I got a Geek from work to help me. It only took about hour. These Geeks are religions about Linux. Hell, I thought the Geek was going to pay me for to put Linux on my used PC!
Reporter: But I thought Lunux was sadly lacking in applications?
Average User: Oh? That's no problem for the "average user", afterall, all I really do with my PC is surf the net and use web based e-mail. Sometimes, when I'm really feeling sophisticated I will crack open a spreadsheet or wordprocessor and maybe even print something or listen to some *legally* downloaded music :-).
Reporter: You really are an average user.
Average user: Yep, and Linux has me covered. But remember I cheap too or I would have spent more money doing the same with M$.