Petreley's column: Make sure you deliver these Linux features soon

This is the week of LinuxWorld Expo, in San Jose, California, hosted by International Data Group (IDG). This latest LinuxWorld Expo is bound to be an exciting show. The March 1999 debut of the show exceeded all expectations in terms of vendor participation and attendance, with the keynote by Linus Torvalds drawing the biggest crowd in IDG show history.

I intend to watch the developments carefully at this and the next LinuxWorld Expo (which takes place in February at the Javits Centre in New York). In particular, I'm looking for a number of events that must occur in the Linux arena before Linux can fully supplant Windows as the leader in enterprise computing platforms.

For starters, one or more of the major Linux distributors or big-name supporters must openly admit that Linux, as good as it is, has some gaping holes in its feature set. I hope to hear a number of announcements at this or the next expo to the effect that these gaps are about to be filled.

Unfortunately for Linux, announcements alone will not do. This is not entirely a fair fight. People seem to be satisfied with shrink-wrapped promises as long as the box has the name "Microsoft" on it. On the other hand, Linux has to actually deliver the features and performance. I estimate that Linux distributors and supporters have about a year to make good on key promises or Linux will lose the crucial momentum it currently enjoys.

Here are some of those key promises I hope to hear.

Enterprise directory services. First, someone needs to lay out a road map for Linux-based enterprise directory services. Ideally, the interface to Linux directory services should be based on Lightweight Directory Access Protocol (LDAP). It doesn't matter as much to me what runs underneath the LDAP interface, but I know a lot of folks who would love to see some combination of Novell Directory Services (NDS) and pluggable authentication modules. Some of the alternatives I've heard suggested, such as NIS or NIS+, simply aren't as robust.

Fail-over clustering. I want to hear a major player endorse an open-source fail-over clustering solution for Linux. Pacific Hi-Tech, the maker of Turbo Linux, is doing some important work in this area. But I would love to see a company with the stature of an IBM announce that it is going to contribute programming talent to a project like this. And I would like to hear someone set a delivery date for a polished, reliable high-availability solution.

Lotus Notes. Linux partners should help Lotus make sure that a full-featured version of Notes ships before February's LinuxWorld Expo. Now please don't go postal on me if you hate Notes. Personally, I am ambivalent about Lotus Notes/Domino. It doesn't take much for me to switch from being a Notes advocate to someone who considers Notes to be a bloated abomination that should be banished from earth.

But regardless of how I feel, the fact is a lot of very large organisations live on Notes. Unless they have the courage to buck the trend, organisations committed to the Intel platform are always going to plan on running Notes on Windows NT. It's next to impossible to make these users take a look at an alternative platform like OS/2. But Linux is a natural because it takes almost no investment to experiment with it.

Imagine, therefore, if NT-centric Notes customers try Linux as a Notes platform. What better way for the NT masses to get an introduction to Linux than by discovering they don't really have to reboot the Notes server once a week to keep things running? You can't make a much better sales pitch than that.

In addition to announcements, I expect the general frame of mind regarding Linux as a desktop OS to change during the next couple of shows. I don't care what the conventional wisdom dictates, Linux is already a killer desktop OS. Since I got Caldera OpenLinux 2.2, I haven't booted up Windows 98 at all except to play games. And as a result I've practically forgotten what crashes are like. I also find myself logging out at the end of a day instead of shutting down. Why bother turning off the machine if you know you never have to reboot to keep things running smoothly?

What events do you feel have to occur for Linux to find a place in your enterprise?

Nicholas Petreley is editorial director of LinuxWorld ( Reach him at, and visit his forum at

Join the newsletter!


Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.

More about CalderaIBM AustraliaIntelMicrosoftNDSNovell

Show Comments