Originally Published: Thursday, 10 June 1999 Author: Andrew Leonard
Published to: columnists/Andrew Leonard Page: 1/1 - [Printable]

Geek Out

A few days ago, I opened up the side pocket of my book bag and discovered I was carrying around not one, not two, but three separate distributions of Linux--Red Hat 5.1 (now hopelessly archaic), Caldera OpenLinux 2.2 (almost too easy to install), and Linux-Mandrake 6.0 (as bleeding edge as you want to be). Man, I thought, I really am getting to be a geek. I've got more source code than I know what to do with....

   Page 1 of 1  

A few days ago, I opened up the side pocket of my book bag and discovered I was carrying around not one, not two, but three separate distributions of Linux--Red Hat 5.1 (now hopelessly archaic), Caldera OpenLinux 2.2 (almost too easy to install), and Linux-Mandrake 6.0 (as bleeding edge as you want to be). Man, I thought, I really am getting to be a geek. I've got more source code than I know what to do with.

No question, Mandrake is pretty neat--proof once again at how fast the Linux community is moving. Basically, Mandrake is a souped-up Red Hat optimized for Pentium processors, with a really nice installation set up that allows users to seamlessly choose between the latest KDE and GNOME desktop releases, as well as other window managers. I was able to install it with lightning speed, though I'm not sure whether that's due to my own increasing comfort with Linux or to the cleverness of Mandrake's coders. (Such is the tragedy of falling in love with Linux--I'm becoming less and less able to relate to my former "stupid user" self.)

In general, judging from Mandrake, news from the Linux front would seem to be good--KDE 1.1.1 and GNOME 1.0.9 are both solid improvements. GNOME, in particular, is considerably less buggy. But as I twiddled around with Enlightenment's themes and kick ass screen savers, I began to wonder just what a development like Mandrake entails for the overall commercial viability of the open source business model.

Given the recent announcement of Red Hat's intent to go public, the question of commercial viability is more relevant than ever. Any newcomer to the free software world is likely to be perplexed at the fact that a vendor like CheapBytes can copy Red Hat Linux wholesale and sell it for whatever price it wants--and it's all perfectly legal. You naturally wonder how Red Hat can make a profit facing that kind of competition. The answer, to some extent, is that Red Hat has a name brand, and promises to back up its products with the kind of support services and documentation that are worth spending money for.

But what happens when a competitor comes along, like Linux-Mandrake, adds a little value here and a little value there to the basic Red Hat package--and, perhaps, also starts selling support services and providing documentation? Their expenses are likely to be much lower than Red Hat's--in effect they are using Red Hat as their own research lab. And after piggy-backing on all of Red Hat's hard work, they add a few deft twists of their own, and end up with a product that is more appealing.

Now, don't get me wrong. I don't necessarily think this is a bad thing, at least certainly not for us end-users. This is why progress in the open source world is so breath-takingly fast to begin with. Everyone gets to benefit from everyone else's work. Onward!

But to be successful, commercially, in this kind of environment, is going to be very, very tricky. Competition will be fantastically intense, and profit margins are bound to be incredibly low. Red Hat's CEO, Bob Young, is frequently quoted as saying that Red Hat's goal is to "lower the value" of the operating system market. In other words, Red Hat has no plans to charge Microsoft-level prices for its operating system, but they'd be quite happy to force Microsoft to lower its prices.

Problem is, Mandrake could say the same thing about Red Hat--which, by the way, significantly raised the price-tag for Red Hat Linux with the 6.0 distribution. How many vendors of Linux--each piggy-backing on each other--will it take before the value of the operating system is effectively zero?

Perhaps this is why Red Hat's SEC filing makes such a big deal of its plans to become a Web portal for Linux--advertising, rather than software sales, may be the company's best bet for profit. But the portal competition is going to be just as intense as the software distribution race.

I love my Linux-Mandrake distribution--it's the flavor of the week for me. But I'll switch in a second if the open source grapevine starts hinting that another distribution has done a better job. And I don't think I'm alone. Why bother with brand name loyalty in the era of open source? The smart decision should be to go with whatever the community as a whole comes up with, piggy back upon piggy back. May the best distribution win!

The question is, after the dust settles, what does the best distribution win? Or does the dust simply never settle?

Andrew Leonard (aleonard@salon.com) is the senior technology correspondent at Salon.com. He is working on a book about free software.

Salon's free software coverage can be found here.





   Page 1 of 1