Paul Murphy wrote a post today explaining what Linux needs to do to beat Microsoft. After some unsubstantiated claims about the skiing ability of Microsoft employees (I prefer snowboards, by the way), Paul closed with this:
Want to fix it? stop trying to make Linux look like Windows, don't put those people in charge, and don't let anyone pretend that Linux is some kind of cheaper Windows replacement. Linux is what it is: Unix, and it takes different reflexes, different ideas about networks, about the role of the computer, about data storage, and about application management to make it work.
I have my opinions about what Linux needs to do to beat Windows. A corollary to that is a point I've made in other posts, which is that Linux would attract more Windows customers by figuring out what those customers like about Windows, and riding that wave into Windows' users homes.
Paul Murphy's proposal, however, essentially says that it doesn't matter what Microsoft does right, we're Unix, dammit, and we gotta be what we gotta be. Why that's a bad idea is best understood by comparing it to other markets.
America's auto industry was, and still is, getting beaten up by Japanese automobile firms. There are two ways to deal with that problem:
A) Argue that Americans cars are what they are, and buyers need to have different ideas about driving and quality if they want to use them.
B) The Japanese are on to something, and it would be useful for us to figure out what that something is so as to copy it in our own automobiles.
I think B) is the best option. Customers are buying Japanese cars for a reason. We could argue that customers are buying Japanese cars for the WRONG reasons, but that isn't very helpful, as it doesn't change the fact that customers want those things.
It needs to be considered that the reason Microsoft is doing so well in servers and desktops is that customers DON'T LIKE the Unix way of doing things. If that's the case, then emphasizing the essential Unixness of Linux isn't a recipe for success. Maybe Apple's approach is better, taking the good things of Unix and wrapping it in a pleasant-tasting candy coating. On the other hand, maybe that's not enough, as Apple hasn't exactly torn the Wintel tapestry from the walls of business.
As I've argued before, the problem with the open source world is they insist that everyone do the equivalent of ditching English in favor of Chinese. Well, instead of trying to pound square Windows pegs into round Unix holes, why not make it easier for those pegs to fit in the first place? That means avoiding the "Linux is what it is" frame of mind, and orienting yourself around what CUSTOMERS want.
If those customers want Windows, ask yourselves: Why is that? Answering that question may require an honest assessment of the relative merits of Windows vs. Linux, but hey, no pain, no gain, right?