In my work as a technical writer, programmer, and consultant, I run into just about every conceivable (as well as inconceivable) notion about this virtual traveling freak show we call a computer industry. That's fine, and it's a major source of entertainment. But one thing that strikes me time and time again is how self-defeating the most ardent Linux supporters can be.

As I see it, there are five truths that everyone in the industry, not just in the Linux community or its more ardent supporters, should accept, for the benefit of themselves and everyone else concerned.

Closed-source software is here to stay. More important, the open vs. closed debate is anything but a black-and-white issue, and it will never be 100 percent resolved in either camp's favor. There's always been open source software, in one form or another, starting decades before a certain college student we could all name decided to write his own version of UNIX. Similarly, there will always be closed-source software, simply because there will always be companies and individuals who perceive, correctly or otherwise, that keeping their source code closed is in their best interest.

Anyone expecting to see the elimination of all closed-source software from the commercial market is just as delusional as the people who think open source is a fad that will disappear quicker than a dot-com's VC funding. The two camps will coexist for a very long time, and they should accept that fact as soon as possible.

Microsoft and Windows, in one form or another, are both here to stay. Microsoft is too big, has too much money and too many dedicated business partners, and is far too wily to disappear in our lifetime. No matter you feel about its products, you can't ignore its market presence or skill in protecting and leveraging it.

Whether Microsoft remains a monolithic company or gets fractured into Baby Bills by the US DoJ is irrelevant to this point. Any hope of reducing Microsoft to merely one of the top 10 software companies is a pipe dream. The only entity that could do that is Microsoft itself, and it shows no signs of suddenly getting an attack of the stupids.

Instead of trying to eliminate Microsoft from the face of the planet, the Linux community should be focused on making Linux the best possible platform in as many market segments as possible. On some fronts Linux is already a wickedly capable competitor that's improving at a startling clip--just look at what it's already doing and is expected to do in the embedded and server markets. The desktop is a whole other mountain, and will take quite a bit longer for Linux to climb, for reasons I'll get to below.

User friendly !=dumbed down. I see and hear this one endlessly, and I honestly can't figure it out. Making Linux easier to install (as many distros have done recently, to a truly impressive degree) doesn't prevent you from tweaking your installation, possibly via using another distro, to your heart's content. Don't like KDE or GNOME or any window manager or even any GUI interface? Don't use them. Maybe you like GUIs, but you don't like some of the programs people are creating that present graphical front ends for configuration tasks that you can do by hand with an editor. No problem--stick to your editor, ignore those utilities, and let the people who need or want the GUI tools use them. No one is taking anything away from you or anyone else with these changes; they're just adding things to Linux distros to make them more capable and accommodating to the needs of other users.

Which reminds me--the people who spout off in newsgroups about any attempt to make Linux more palatable to "dumb lusers" are only helping Microsoft and hurting Linux. Right now, Linux is in a precarious state in the desktop market. The number of potential Linux users, almost all of whom are computer literate Windows users who've heard about Linux and are curious about it, is probably far greater than the number of people who currently use Linux as a desktop OS. When people mouth off about not wanting to see Linux "dumbed down," it casts Linux and the whole community in a bad light, and makes it that much harder to win converts.

You can't fight Microsoft and not cater to Windows users. It's this simple: Windows has 90 percent of the desktop market. If you want Linux to succeed on the desktop and present a genuine alternative to Windows, you have to actively court those people. If you think it's not Linux's role to challenge Windows on the desktop, then I can only assume you want Microsoft and Windows to own the desktop for the foreseeable future, since Linux is the only challenger in sight.

The core issue is how to cater to those users without turning Linux into OpenWindows. I'm not suggesting in any way that we clone Windows. But I do think we need to focus on the barriers to entry that are keeping Windows users from trying out Linux, and figure out ways to lower or eliminate those barriers. That requires us all to think like those users and figure out ways to give them what they need and want without compromising Linux's traditional strengths. In some cases that means borrowing ideas from Windows (or the Mac OS or OS/2 or wherever), sometimes it means learning from other OSes' mistakes and not doing certain things, and yet other times it means reinventing a major piece of Linux or one of its components.

I believe we all have to decide if we want to help Linux compete against Windows for the mainstream desktop, or let it remain a desktop OS and plaything for the Übergeeks. I think unseating an entrenched power in the mainstream market is as big (and as much fun) as challenges get in this business, and I can't wait to see Linux do it.

There are really three meanings of "free." The Linux community often observes that there's free as in beer (zero monetary cost) and as in speech (legally unfettered). But there's also a third relevant spin on "free": "Free" as in "the most expensive thing in the world is free sex." (Remember Fatal Attraction, the movie that scared even perfectly loyal married men spitless?)

The "free sex" meaning is paramount among MIS managers, and it explains why they're not jumping for joy at being able to outfit thousands of PCs with Linux with a single $1.99 Linux CD from cheapbytes.com--they know that acquisition cost is only the tip of the tip of the iceberg; the real cost is the "hidden" details, like user training, interoperability, and maintenance. This is why Linux is exploding in servers and embedded devices, where these other costs aren't a factor. Companies can use Samba to make Linux impersonate a Windows NT server, thereby insulating their users from the change. They can also pick up the source code to Linux, change it as needed, and then use it at a marginal cost per manufactured unit of $0 in embedded devices, where, again, users won't see it.

On the desktop segment, the after-the-fact costs typically swamp out the acquisition cost of an operating system. The fact that Linux, thanks to its robustness and efficiency, can often lower those other costs is moot, for two reasons: First, the decision makers don't have enough evidence to bet their careers on such a massive change; Second, many of them still don't know about it. To a large part, success in the business desktop will be a matter of educating the right people about Linux's current strengths, not changing the OS.

I'm optimistic, though. Thanks to the relentless efforts of many projects, companies, and individuals, Linux will succeed on the desktop eventually, but it will be a slow, evolutionary process, not a revolution. We just all have to keep our eyes on the prize, and pull in the same direction.

About the Author
Lou Grinzo is a freelance writer, programmer, and consultant specializing in Linux. He can be reached at lou@gizmodrome.com.