|
Author: Con Zymaris, <conz@cybersource.com.au> |
Modificatiion Date: |
Modification Number: 20 |
I recently came across a pamphlet prepared by Microsoft and aimed at corporate and government Information Technology executives.
It is obviously Microsoft's business to prepare and disseminate material which bolsters their product marketing message. However, I deem it necessary to prepare a response, in kind, when there are obvious mis-truths and deceptive statements in any such message.
I make the pamphlet in question available for review here: http://demo.heins.net/comp/images/MS_vs_Linux.png (Thanks to Mike Heins.)
(A slower link mirror is here: http://www.cybersource.com.au/users/conz/MS_vs_Linux.png)
Unfortunately, it's a scanned image of a photocopy, so the quality is not great, but the text of the message is legible.
Let us now work through each of these points that Microsoft raises, and analyse each of their comments about Linux and Windows.
Under the heading of Better Productivity, Microsoft state:
"The independent research firm AIR found Windows users accomplished tasks in half the time of Linux users."
Analysis:
A copy of this AIR study cannot easily be found on the Internet and Microsoft does not provide a link. However, we do have another more recent and openly published report by independent firm Relevantive (source: http://www.relevantive.de/Linux_e.html) which looks at the usability differences between Linux and Windows XP. To quote the PC World report:
"The researchers studied how easily two groups of users could perform tasks using the different operating systems. One group consisted of 60 users between 25 and 55 with computer skills but no prior experience with Linux or Windows XP. They tested the pre-configured open source software according to various criteria, such as the ease of creating and administrating new and existing files, copying CDs, and performing some basic office tasks, such as composing a text and sending an e-mail."
(Source: http://www.pcworld.com/news/article/0,aid,111871,00.asp)
Also, the study calculated that the time difference to undertake these sets of activities between the two platforms, only differed by on 8%. Finally, the report conculded with:
"80 percent of the Linux users believed that they needed only one week to become as competent with the new system as with their existing (presumably Windows -- Ed) one"
Under the heading of Lower Cost, Microsoft state:
The technology firm IDC found Microsoft Windows & Linux servers cost about the same over 5 years.
Gartner, the leading IT industry analyst firm, found, with higher deployment and management and training costs, the total cost of Linux software adds up to and can even surpass those of Windows.
Analysis:
The author of the report that Microsoft mentions, Dan Kusnetzky, has made it clear that this report that Microsoft commissioned (and paid for) was designed to show Linux in a less than favourable light. The following appeared in a recent BusinessWeek article:
Yet even this tactic seems to be backfiring. One of the study's authors accuses Microsoft of stacking the deck. IDC analyst Dan Kusnetzky says the company selected scenarios that would inevitably be more costly using Linux. Also, he believes Windows should be cheaper to operate, since it has been around longer, giving Microsoft more time to develop software to manage the operating system. "Microsoft has had a lot more time to work on this. I wonder why the win wasn't bigger," Kusnetzky says.
(Source: http://www.businessweek.com/magazine/content/03_09/b3822610_tc102.htm)
Recently, Gartner published the following Total Cost of Ownership study, comparing locked down Linux and Windows XP:
"For knowledge workers, locked Linux is $5,062. For WinXP, this figure is $5,908."
(Source: http://www.newsfactor.com/perl/story/22012.html)
Thus, according the Gartner, Linux thus has a 15% lower Total Cost of Ownership than Microsoft's flagship corporate PC platform.
Under the heading of Better Security, Microsoft state:
Windows received the highest level yet on the Common Criteria Certification, an independent auditing program for technology and THE standard measure for security recognised by governments worldwide.
Analysis:
Microsoft further state that no Linux system has undergone or passed such rigorous security testing. This is incorrect.
LINUX software has been approved for use on the most sensitive computers inside US corporations and government, including those inside banks and the Pentagon.
(Source: http://australianit.news.com.au/articles/0,7204,6869218%5E15321%5E%5Enbv%5E15306,00.html)
Furthermore, Microsoft's level of security flaws, in comparison to the level of functionality components bundled with the operating system, is perhaps 20 times higher than the corresponding Linux and Open Source level. This can be adduced from the following simple concept. When Microsoft ships a professional Windows platform, like Windows 2000 Server, they provide around 120 functional application systems. When a Linux vendor, like Mandrake, Red Hat, or Suse ships a corresponding professional platform, they supply over 2600 functional application systems.
An analogy can be made here. One business-man turns and asks another how many accidents he has for his fleet of cars each year. “On average, 5 accidents per year” is the response. To which the first business-man retorts “Well, we only had 3!” The second business-man asks “Yes, but how many fleet cars do you have?” The first replies proudly “10 cars.” The second then smiles and says “We operate over 200.”
Whilst this is a pithy example, the point is made. One cannot judge security based on the count of vulnerability alerts alone. Therefore, even if there are more security patches issued to Linux, the application base from which the applications are drawn is substantially bigger than for Microsoft's software offerings for Windows. (Source: http://www.cybersource.com.au/users/conz/linux_vs_windows_security_alert_comparison.html)
Under the heading of Better Interoperability, Microsoft state:
More applications work with the Windows platform than any other, including Linux.
Microsoft further states that 'In many cases, you'll have to write your own driver to get an application to work with Linux software.'
Analysis:
Firstly, Microsoft mentions 'interoperability', but what they really mean is 'availability of applications.'. Linux is widely acknowledged as the most interoperable of all computing platforms, working well with Novell, Apple, DEC, IBM, Sun, HP, Microsoft and many other protocols, file formats and data sets. Historically, Linux has had to be interoperable in order to successfully permeate and blend-into pre-existing networks and environments. Also, Linux's user-driven nature helps Linux remain interoperable with most other computing platforms. This is in contrast to the often Not-Invented-Here mindset of many vendors not supporting technology produced by other vendors, or altering that technology somewhat to break interoperability to strategically lock-in users to that vendor's platform.
No one can dispute that there are many more applications available for Windows than there are for Linux. However, this needs to be analysed. Windows has been a desktop operating system for over 18 years, and on the server for perhaps 10 years. Linux has seen uptake on the desktop for perhaps five years and on the server for perhaps 10 years. As such, Linux has more 'server oriented' software than Windows, and Windows has more 'desktop oriented' software than Linux.
Regardless, it can be said that both platforms have the full panoply of applications that most organisations need to run.
The statement that Microsoft makes about users needing to write their own drivers to get applications working with Linux is absolutely false. Sometimes Linux developers take on the task of writing drivers to work with esoteric hardware. As Linux has become mainstream on the server, so too is it automatic for server hardware vendors to automatically supply Linux versions of their drivers, much like they do for Windows. Linux is now becoming mainstream on the desktop, so hardware vendors will also supply drivers for desktop-oriented hardware too. For example, Linux supplier Red Hat lists drivers supporting products from over 400 hardware vendors. (Source: http://hardware.redhat.com/hcl/)
