Do me a favor, don't try Linux

Recently, there seems to be an abundance of articles on failed Linux evaluations in corporate environments. Most of them point out why Linux didn't make the grade for one reason or another. As Linux becomes more of a viable option for desktop deployment, I suspect we will see more of these types of articles. I, however, am not too sure they are all that enlightening.

Are you serious?

I begin to question some of these attempts at getting a fair look at alternatives. If the entity is serious about evaluating a Linux desktop what kind of testing program are they implementing? Two recent articles detail single user tests, not a very good survey if you ask me. Most also seem to be heavy Microsoft houses, using Exchange/Outlook, IIS, Microsoft SQL Server. There's nothing wrong with this as long as you concede that you are tied in to Microsoft technology. Don't expect a client replacement for Outlook that utilizes many of Exchanges features. There isn't a Windows client replacement for Outlook, let alone a Linux one; however, there are alternatives depending on your organizational needs.

Performing a limited test of an alternate desktop does little else than give you the false impression that your current choice was the correct one. Giving one user with one role and making it a test case gives you an accurate view of one user with one role. This is hardly a method of determining the downsides and benefits of what you are testing.

Smooth as a cats tongue.

Any test will have its ups and downs. Hardware changes, drivers, software changes, drivers, training, drivers, deployment, and of course the problems drivers present. I am always bewildered with arguments of drivers and the limited hardware support of Linux. The alternative gives the wrong impression that all the drivers for every piece of hardware in the organization is on that Windows CD from Microsoft, or on that the driver disk that came with your system. Apparently no one has ever had to download the latest drivers from ahardware manufacturer. If they did have to download a driver, for some bizarre reason, they magically worked every time. How convenient for management not to know what IT has to go through with each upgrade, update or new piece of hardware.

I also find it interesting that a department will purchase a laptop with Windows XP or OS X pre-installed, yet do not do the same for Linux. There are vendors that sell pre-installed Linux with all of the driver issues resolved. Then you can evaluate the operating systems on functionality, stability and security.

Those three little words.

Nearly every Operating System vendor has a Hardware Compatibility List (HCL). If you decide to run a stable enterprise operating system, take a look at the HCL to see if the hardware is supported. If the hardware is not supported and there is no alternative, look for another distribution, one that meets you specific needs. You may also find that your old hardware is well supported and there may be less of a need to purchase new hardware.

Imitation is the best form of flattery.

Many organizations do not like to take the lead in implementing technology. Many will learn from the mistakes of others and base decisions on others implementations. This is where these types of articles will fail the industry. These wait and see organizations will likely fall into the "we made the right decision" trap based on limited and inconclusive testing. Large organizations will have many types of user roles and implementing tests on those users, who will benefit from alternatives, need to be determined. To those organizations who insist on giving results of, insincere at best, tests do me a favor and don't use Linux. I'm sure you had no intention to anyway.


Verbatim copying and distribution of this entire article are permitted worldwide, without royalty, in any medium, provided this notice is preserved.