Linux needs to play its own game and quit comparing itself to Windows by D Colbert

I'm fairly frequently accused of being a Microsoft Fanboy or Corporate Stooge for what it seems many Linux fans mistake as an "anti-Linux” philosophy and outlook. In my own defense, I don't think I am "anti-Linux" I think the main problem is that the Linux community pictures itself as a looming threat directly to Microsoft "dominance".

Frankly, Linux spends too much time directly trying to compare itself to Microsoft in areas where Microsoft dominance is virtually insurmountable, and where Linux, despite significant advances, finds itself (relatively) lacking. If I were a politician, I wouldn't campaign by claiming:

"My opponent claims that I am inferior to him in foreign policy, and while this may be true, I have made SIGNIFICANT advances in my foreign policy skills. Certainly my opponent is the BETTER choice for foreign policy, but hey, I'm a NICER guy, and even if I'm not as good, I'm much better than I used to be".

Unfortunately, that is exactly what Linux does when it tries to compete directly with Windows on the desktop OS platform in areas like Plug-and-Play support for the widest variety of consumer accessories. I think it stings the Linux community to hear things like this put so bluntly.

I also think the Linux community has a natural desire to avoid facing these facts, and that as a community, it would often rather embrace a fantasy world where Ubuntu is really a compelling alternative to a Windows OS that could threaten to disrupt Microsoft dominance. Speaking the truth in the Church of Linux is an excellent way to get ex-communicated. But then again, people have been running into trouble by speaking the truth in religious organizations for centuries.

And therein lays another problem. I don't think the Linux community lies, but I don't think they make accurate or fair comparisons. Whenever I hear figures on the number of datacenters with huge Linux deployments, I always wonder where all of these data-centers are.

With 15 years of experience, I've seen Sun Solaris heavy data centers, I've seen HP-UX heavy data centers, and I've seen LOTS of big corporations with huge Windows deployments. But I've seen very little Linux, scattered here and there, and often in strange, supporting network and infrastructure roles.

Another (ironic) example I use is that when you see a giant LCD on the Vegas strip that has crashed, it isn't LINUX you see underneath it. When you're cable TV guide has crashed, it isn't Linux. When you're in the Build a Bear workshop and a PC is down, it isn't Gnome that the app crashes back to. And that isn't about Linux versus Windows reliability - it's about applications, and we all know Linux apps crash as frequently (or more than, in the case of KDE) as Windows apps.

Some people claim that I have conformational bias - that I see more Windows because I work in a Windows-biased segment of the industry. But I think it is clear, more than 9 out of 10 PCs you run into in the private and public sector outside of very narrow niche industries are going to be Windows-based. The number of Linux machines becomes grossly inflated through several methods, many of which do not compete with Microsoft technologies and actually assist Microsoft while keeping Microsoft's real competitors tied up worrying about additional competition.

For example, I've got a feeling that the Linux numbers we see count every embedded Linux device - consumer or otherwise - available on the market. I think the Linux community takes these artificially inflated numbers and compares them to Microsoft's Windows desktop and data center numbers. I think that Linux numbers we see count every download by every curious user who may or may not ever get around to installing their download on an actual physical machine.

Most importantly, once we get down to the nuts and bolts of it, I think Microsoft does very little competition on the infrastructure side of the enterprise, such as DNS, Active Directory, DHCP, and Web/IIS. You might also count FTP and e-mail.

It occurs to me that Microsoft really doesn't care when Linux "grows” enterprise market share, because it is the OS under the interface of the new IBM XIV storage solution. Effectively, Linux is an appliance in this role, and the LINUX is not as important as the role of the appliance.

Often, the Linux is so transparent that as a data-center administrator, you may not even know that Linux is in your data center. There is no conceivable way this is a threat to Microsoft. It isn't even really fair to use those numbers when comparing Windows installs to Linux installs. I said to a friend, regarding the Tom-Tom lawsuit,

"Linux is very popular, as long as no one needs to know it is there”.

It's great that Linux can have this space and compete with products like Win CE/Mobile/Phone - or be in the enterprise providing supporting infrastructure roles that Microsoft doesn't want to compete with in the first place. Linux becomes a foundation on which companies transparently build other turn-key products as their core business for retail consumption.

Windows can be used in this role, but is generally used in a far different business model - as a component of productivity-enhancing business machine solutions running a variety of off-the-shelf, user-selectable applications. But when the Linux community speaks of "competing" with Windows, they're really talking about this second model - and it isn't, it absolutely isn't, the place where Linux performs the strongest. This is just plain silly. Who ignores what they do better while trying to compete on what they know they do worse? Republican candidates, in the last election, that is who.

I have Linux in my Windows environment. Cisco Call Center Manager for VoIP is a Linux-based utility. I'm in very critical discussions right now that will likely result in a migration from EMC to IBM network storage solutions - and that solution is Linux-based. In both cases, I do not need to know how to compile kernels, how to grep, how to use package managers, how to use VI to edit config files in the /etc directory. As a matter of fact, with the IBM XIV solution, I couldn't do that if I wanted to.

These solutions do not displace a single Windows solution in my environment. They facilitate better delivery of my Windows-based solutions and improve the experience for my end users. The IBM XIV solution is called disruptive technology, and it may very well be.

Linux is the foundation of a disruptive technology that threatens a current giant in the technology marketplace. Unfortunately, it isn't Microsoft, it is EMC - and I imagine that Microsoft has no strong feelings one way or the other about that. It is certainly a huge win for Linux, and it illustrates that Linux has a viable place in the technology sector. What it does not do is show that Linux is, or ever will be, a threat to Microsoft dominance of the back office, plus corporate and end-user desktop.

Ultimately, I am not married to Microsoft technologies. I support Microsoft because that is where the significant amount of business demand is. In my own data center, I'll deploy the best, least expensive, most reliable solution in every case, regardless of what OS platform underlies that solution.

So, I am certainly not vocally anti-Linux. I'm opposed to forcing a square peg into a round hole because you think the square peg is actually the superior round peg, or because you think the round peg maker isn't a very nice businessman, or because you think that the philosophy of the square-peg community is more forward-thinking than the philosophy of the round peg community.

I don't like it when musicians and actors preach science and political philosophy to me. I don't want my business solutions to be based on making productivity sacrifices that make me less able to compete.

Ultimately, I think this is what Linux versus Windows debates come down to - are you open source and willing to sacrifice in order to support open source, or are you in business to make a profit and be the most competitive force possible in your market? Political, ideological, and philosophical differences are at issue here, because clearly, Windows solutions dominate and are superior for desktop, back office, and application-hosting solutions on the enterprise.