How Linux Could Kill Windows NT

How Linux Could Kill Windows NT

Jesse Berst, Editorial Director
ZDNet AnchorDesk

June 23, 1998

Linux. It's not a serious challenger to Microsoft Windows NT. But it could be. If three things take place.

Linux is a freeware operating system developed in the early 90s by a volunteer group under the leadership of Linus Torvalds, then a Finnish computer science student. This 32-bit, UNIX-like, multi-user, multi-tasking operating system is legendary for its stability. And for its flexibility, since users have full access to the source code.

Linux has roughly 5 million users worldwide, compared to about 200 million for all versions of Windows. Click for full story [ http://www.zdnet.com/zdnn/content/inwk/0518/315679.html ]. Lately I've been seeing signs of momentum.

Corel is shipping Linux as the operating system on its NetWinder network computer. Click for full story [ http://www.zdnet.com/pcweek/news/0504/08elinux.html ].
Sun Microsystems is backing Linux as an alternative operating system for its UltraSPARC platform. Click for full story [ http://www.zdnet.com/sr/breaking/980601/980605e.html ].
VARs and system integrators have begun to use Linux for customer projects. They like the fact they control the source code, so they can mix and match the components they need and build custom extensions.

But these are baby steps. Three things are needed if Linux is to duel Windows NT for real.

1. Enterprise-quality technical support. Linux still feels risky to large corporations, who feel they need support from a single point of contact, not from a loose alliance of libertarian programmers. It makes some IT professionals wonder if they could get fired for choosing Linux [ http://www.zdnet.com/anchordesk/story/story_1774.html ].

Of the three obstacles, this is the easiest one to solve, because it is largely perception versus reality. Companies such as Caldera and Red Hat now offer user-friendly commercial versions with nationwide support. Click for full story [ http://www.zdnet.com/products/content/pcmf/1231/267877.html ].

2. Tier One applications. The leading applications must be available in Linux versions. The situation is improving but very slowly. Corel ships WordPerfect for Linux and plans to develop a suite of business applications. Most other vendors won't do Linux versions until there's enough demand. And there won't be demand until there are enough applications. That leaves Linux trapped between a chicken and an egg.

3. A standard interface. This issue will prove hardest of all because it flies in the face of the Linux gestalt. It's not a technical issue. At least one firm has already built a Win95 clone. Rather it's the problem of getting the fiercely independent Linuxites to agree to a single standard. Remember, the core community is made up of Unix geeks who think graphical interfaces are for sissies.

Would you like to see the rug pulled out from under Microsoft? Here's how it could happen. IBM [ http://www.anchordesk.com/glossary/glossary_158.html ] ships and supports Linux. Oracle does Linux versions of all its products. A consortium of top vendors picks a standard Linux interface and creates a compatibility logo.

Possible? Absolutely. Microsoft's long-range strategy would come crumbling down if it was forced to give away Windows NT. It plans to upgrade all of us to NT, and charge us twice what it gets for NT. Is a Linux takeover likely? Give me a break. Of course not.

Still, as we reported yesterday, IBM is now shipping the freeware Apache Web server. Is it too much to hope that it might ship and support Linux? Click the TalkBack button to tell me what you think. Or jump over to the discussion now underway in my Berst Alerts forum.