According to industry analyst firm Gartner, the technology industry has entered a period of high commercialization of Linux. Many acknowledge that Linux may have a place in their organizations, but many questions remain – questions about the business and technical merits of Linux, what risks will need to be managed, and its future direction and roadmap.
It was with that in mind that Microsoft Canada engaged Compugen to review Windows Server 2003 and enterprise-level distributions of Linux from Red Hat and SuSE to help understand the practical implications of using one operating system versus another. The objective was simple - to produce a reference document to help IT managers and professionals consider their options. The test compared the usability characteristics of the three operating systems when it comes to performing a set of typical IT tasks from setting up user profiles to installing to managing desktop security.
Key findings from this test and demonstrations of how each platform fared in solving these IT tasks were given to the press last week. Liam Lahey from eChannelLine Canada was one of the journalists in attendance last week and has already published his article on the event. http://www.integratedmar.com/ECL.cfm?item=DLY041005-5
The comment I think is most relevant is “Windows Server 2003 provided better functionality and usability than SuSE Enterprise 9 and Red Hat ES 3.4 primarily due to the immaturity of the Linux system management tools and a lack of supporting documentation.”
I think what companies need to remember as they look at open source software, is the level of support and accountability that they can expect to receive, and who they expect to receive it from.
Read the full report on the Compugen site.
What are your thoughts?
UPDATE: (May 6, 2005)
The Web cast is up and running from this event. Head over to http://www.insinc.com/hrc/20050508
If you listen to Gartner, you're not doing the due dilligence to do your job.
As long as Microsoft is paying for studies like this, the objectivity of the results will always be doubted. I almost wonder why they bother...
Linux has never improved. GUI's all still suck, there's no software, ease of use is non-existant, drivers are non-existant. Games don't barely exist for it...Linux is zip. The only good thing you will hear about Linux is from the wackos who are so blinded by reality they continue to use it because of their hate-hardon for Bill Gates. Is linux less hack free. No. Who bothers to smear Linux hacks in public media streams? It's rare to see but it occurs and Linux has many exploits and many that people don't even know about. Less users, less visibility. Like it or not Windows rules Linux world.
Just read the "study" -- it's mostly a bake off between two different admins on different platforms. Not even counting the skewed criteria, it doesn't pass first muster on any scientific basis for any conclusion stronger than "this MS admin had a better time using Server 2K3 than this admin had using RH and SUSE".
Nothing worth hyping or hyperventilating about ...
I've only read the first half of the report carefully and skimed the rest, really just looking for your comparative comments between Windows and Linux.
In many comparisons where Linux had equivalent functionality you fail to consider many great solutions available to Linux. For instance, Remote Server Management can easily match 'Remote Desktop' when VNC or XWindows itself is used as the interface to the remote machine!
The fact that you state over and over again that the Linux tester required "more expertise than Windows" is extremely strange... What does "more expertise" really mean? Does it mean that an admin is expected to know the formats for DNS zone files along with the Apache config file rather than relying on wizards? Ok, then yes, it's expected that the admin be trained and have at least a mimimal amount of retention for these things. For the record I'm a Java Architect and Linux server hobbiest and seem to have no problems mustering this level of expertise. It's comical to use this phase in your comparisons.
As far as a near universal configuration tool for Linux is concerned, it's call VI. Excellent alternatives include emacs, vim or pico for those with a passion for control keys.
Finally, if you are to compare operating systems, please actually compare features in common to each. Shadow Copy, for instance, seems like a feature much more germain to Windows systems than to Linux. There are countless features found only in Linux (support for dumb terminals) that are poorly supported in Windows, yet don't make it into your report. Please keep the comparisons pertinent and fair in the future.
Linux did not win in ANY category? Does anyone find that difficult to believe? Further, this "test" sought to compare Windows-specific features with things that don't even exist in the Linux world. Instead of comparing apples to apples, they are comparing apples to null. Guess who wins?
The only salient point here is the less-than-integrated nature of Linux configuration files. I can agree with that, but its not something that cannot be overcome.
Linux forces you to deeply understand the workings of your hardware and services. And that is a GOOD THING, because it makes for a more informed, better-equipped administrator. HIding everything behind a pretty GUI may make initial setup easy, but wait until something goes wrong. The GUI becomes a handicap.
Barnaby Jeans, Microsoft Canada National IT Pro Advisor
That tells it all.