So it is that I was recently struck by two stories that showed me once again that while many open-source people are technically brilliant, they can be business idiots.
In the first story, ace reporter Stephen Shankland reported on the problem that Red Hat faces because of companies that take RHEL (Red Hat Enterprise Linux) source code and turn it into their own products. He quotes several people who are overjoyed at being able to use this code without paying Red Hat one thin dime for its distribution.
Those people are missing the point: Red Hat is not in the Linux-distribution business. It is in the Linux-distribution service business.
Red Hat doesnt need the business of people who can take source code and run it successfully on their desktop systems. If that were the case, it wouldnt have taken the bold move of dropping its Red Hat Linux line and shifting all the companys focus to RHEL.
Indeed, Red Hat very deliberately abandoned this box-oriented approach. When I talked to Red Hat executives last year, they told me they already know that theres not enough cash flow to be found in trying to marry the old proprietary box-software business model with open-source software to make a viable business.
Businesses buy into Red Hat, or Novell/SuSE, because the companies deliver support and service, not because they are now running the latest Linux 2.6.11 kernel. Support and service is how successful open-support businesses make their money.
And then there is the second story, regarding Mozilla.
The good people at the Mozilla Foundation can make great software, but boy, can they be clueless about getting corporate America to trust them.
First, there was the problem of timely updates, which Mozilla seems to have gotten a lot better at in a hurry with the recent release of Firefox 1.02. But theres another, deeper problem.
In a recent Slashdot interview, my friend Robin "Roblimo" Miller asked Mozilla Chief Lizard Wrangler, aka Project Director, Mitchell Baker, what the Foundations plan was for Mozilla support now that its retiring the popular Web and e-mail package.
She replied, "So that ones been two years. 1.7 is likely to be in that range. And by supported, we mean maintenance and security releases as needed."
Some open-source supporters put their careers on the line to get their companies to adopt Mozilla and now they have to face their CIOs and, worse still, their chief financial officers, with the news that theyre going to have to switch end-user Web packages someday soon.
Let me tell you, thats going to go over really big.
Yes, I think that Firefox and Thunderbird are better too. So what?
Businesses arent about whats the best program right now. Theyre about stable, supportable programs that they know will run for at least three years without any major upgrade or training worries.
I couldnt blame a CIO who had just been burned by Mozilla for moving not to its superior other offering, but back to Internet Explorer instead.
In fairness, Baker also said, "We continue to evaluate whos using things, how many people are out there, what do they really need." But that kind of fuzzy talk doesnt make businesses feel warm and cozy.
Look at Novell in comparison. Novell has committed to supporting its newly open-sourced Hula (a lightweight collaboration server—think an open-source Exchange) for at least five years, while its also committed to supporting its heavyweight collaboration server, GroupWise, for at least ten years.
Yes, Mozillas source code is open, so users could, and will, support it themselves. But the last thing most businesses want to do is maintain code or rely upon the hit-or-miss support of mailing lists or fan-based Web sites.
As the saying goes in IT circles, what a business wants is one neck to choke: a single partner to depend on for its support needs so that it can focus on what it does best instead of wasting its time maintaining code.
So, do you want to know another way to make money at open source today? I think, for a few years, anyway, you could do well by offering support for Mozilla. After all, Progeny Linux Systems is doing just fine by supporting the earlier revs of Red Hat.
If open source really wants to gets out of the back room and into the office, open-source supporters need to support open-source projects for the long run. Thats really what businesses want: programs that not only work well, but will still be fully supported five years down the road.
It really is that simple.