Editor's Note: How Not to Define LinuxAug 19, 2005, 23:30 (18 Talkback[s])
(Other stories by Brian Proffitt)
By Brian Proffitt
A friend of mine asked me if Linux was getting too commoditized and too ordinary. A fellow Linux user, he seemed worried about the "wow" factor of Linux dropping below the point where it would capture any interest.
As the editor of this site, my immediate response was to frame my answer in terms of Linux Today. Basically, I told him, I wasn't concerned about the lack of interesting news regarding Linux. Linux is far too diverse and broad of a platform to let news about it fall away to nothing. I predicted that other proprietary companies will become less and less interesting in a news-sense as time goes on, because their news is controlled by single entities. People recognize that more and more lately, and the desire for independent news sources is still growing.
Later in the week, I realized that I may not have answered my friend's question in the manner that he wanted. He mentioned a decline in interest in Linux, and my knee-jerk reaction was to couch it in publishing terms.
If you have taken philosophy or sociology or have just hung around bars for a goodly length of time, you will know that this is nothing new. We all tend to frame the world around us in our own terms, locked away inside of our skulls day in and day out. But it tweaked an idea in me that I thought I would share.
I think total cost of ownership studies are stupid.
Even if you could make up a TCO study that was completely fair and impartial, the results would be meaningless. Not just on a personal scale, but also on a global scale.
Why? Because anyone you present the results to are going to be looking for different things and will interpret the results in a different way.
Now, bam! Right out of the gate, the whole lot of Windows supporters are going to come out and say "Brian Proffitt is giving up on TCOs because they don't tend to favor Linux."
No. I am rejecting TCOs because they never will cover the hundreds of different variables that make up one company's unique business needs. And that includes, by the way, the TCO studies that do favor Linux. I simply think they are all unsupportable because for one, everyone will read exactly what they want to read out of them and for two, they don't address the nuances of what a migration to Linux can do for a company.
Here's a recent example: my daughter was helping me break down some old office furniture this week, and was looking for an Allen wrench to remove some screws holding my desk together. It took her a while to find it, and then when she did, she laboriously began to twist the screws out one at a time. While I am all for teaching my kids the value of a good honest labor, there is such a thing as doing too much work for a task. Calmly telling her to step out of the way, I began to whack the desk components apart with a rubber mallet. Inside of two minutes, the desk was in pieces.
I explained to her that since this desk was falling apart and was bound for the trash, there was no need to carefully disassemble it. In this case, breaking it apart in a controlled manner was justified.
I see a similar situation when it comes to businesses and Windows. Because of the way Windows is designed a lot of company decisions were (and still are) made based on the limitations and capabilities of Windows. The business is set up this way because that's what the proprietary tools will let you do. Just as my daughter expected to have to use the Allen wrench and carefully disassemble the desk, so too do managers and decision makers expect that all IT tasks have to be approached in the same way with the same tools every time.
Switching to Linux, though, means there is much greater flexibility in how things are done. You can still do things in Windows' terms, or you could be more original. If I want my sales team to remotely input their sales straight into the accounting system, I could set up a way to let them e-mail the numbers in, then build a macro or application to parse the data and put it into the accounting software, using existing tools. Of course, if I do this in Window-land, I have to be aware of security and stability issues in these existing tools and hope my data doesn't get discovered. If I am using Linux, these issues are nonexistent. Plus, I could likely just bypass all of these daisy-chained tools and build a smaller faster client app that will hook directly into the accounting system and do the job more efficiently.
Like taking a hammer to an old desk.
This is the biggest thing TCOs miss. They assume that if you switch to Linux, then you will continue to do things in terms of the ousted Windows system. Maybe you will, because inertia is a hard thing to overcome. And, if you do, you may indeed come up with higher TCO numbers for Linux over Windows. Because (and here's the big point):
Using and judging Linux in terms of Windows is completely wrong. Linux is not Windows and Windows is not Linux.
Yes, there are many capabilities that overlap between the two operating systems. But there are many that are unique to each operating system. The problem is, analysts and media tend to look at Linux in terms of Windows and judge it accordingly. They also tend to look for the negatives, because even though they don't say it, their group think is that Linux is still a precocious upstart.
Here's an example: "Linux doesn't have enough applications on the desktop," they say. Well, where's the "enough" coming from? Enough for a home user? Enough for a business user? Based on what criteria? Press them, and they will say "the Windows desktop."
But do they ever say, "Windows does not have the flexibility to truly function well in a business environment?" No, they don't, because in the mindset of the Windows user businesses use Windows every day and therefore it is a proven fact that Windows does well in a business setting. No, businesses have had to bend over backward to accommodate the inflexible nature of Windows all of these years, and they honestly expect to have to do the same with Linux.
With that mindset, why would they switch?
Microsoft is well aware of this. Because they read the TCO studies and the market research papers too. And they have seen that while they still hold a vast majority of the world's desktops, there is a lot of grumbling going on about business rules and IT management. And they see where Linux will beat them time and time again: it's open, and so are many of the apps that run on them.
This means getting Linux to accommodate a business rather than the other way around is a much, much easier proposition than Windows.
That's the kind of perception that gets missed in TCOs. And that's the kind of thinking that Microsoft doesn't want people to hear. As long as they can hold comparisons in terms of their products, Microsoft will be happy. But the day is coming when people will start to think of IT in terms of what they need to do, not what they have to use. The recent success of Firefox is a good example: Firefox succeeds because it sets a new standard for browsing. It does tasks on its own terms.
Who cares about Windows? Linux can be defined in its own terms and will succeed quite well on its own.