Robert Pogson

In the “Good Old Days”^TM, I remember the times when we “processed” chickens back on the farm. Chickens, like many intelligent life-forms, are not programmed to die and serve themselves up on our dinner-plates. To get them to cooperate in their executions was a bit of a chore. We held the bird up on a chopping block and let go the axe… Doing that with a struggling bird was not an exact science.

I remember my father telling me about his “Good Old Days” when they used to put the chicken’s head down on a white line and the chicken would mesmerize itself staring down the line and be still while the axe was put to use. He, himself, did not use that technique so I cannot really say whether that approach worked.

It seems Wintel wants that technique to work on the world of IT to change from XP to “7″. Windows 7 Migrations: Don’t Get Distracted by XP, Windows 8 is the title of an article that claims the right thing to do is to stay on the Wintel treadmill for another round and forget that tablets exist and will be a major part of IT. The fact that there is no benefit at all to the “chicken” is beside the point. If someone has XP working well enough and it will be supported until 2014 it is not logical to conclude that a change to “7″ is the thing to do in the next few months.

There might be some performance/reliability improvements in migrating to “7″ but businesses are much further ahead to migrate to thin clients no matter what OS they use, because it is a quick exit from the Wintel treadmill. It is insane to change the OS on millions of clients every time M$ belches another release. Car makers develop new/improved models annually and we don’t change our cars annually. M$ could make a new release every year or every decade and it would not be a reason to change. Businesses are far ahead to put the OS of their choice on the client and do everything on a terminal server or cluster rather than working for M$’s bottom line.

I recommend businesses use GNU/Linux on their thin clients. There are tiny distros made for the purpose and even a full-bore distro like Debian GNU/Linux will work well in the role. Businesses get the advantages of lower cost and more reliable client machines and higher performance on the terminal server whose hardware they can upgrade as/when they see fit and if they are truly locked-in to M$ they can change the OS on the terminal server when M$ releases. That’s the right way to do IT. Keep things modular and the modules independent so you can change components when the time is right without messing up anything else. Putting most of the applications on the terminal server or web server makes it possible to get off the Wintel treadmill and onto your own path to profitability.