Tempest in a Caldera
But there has been a definite decline in decorum lately. The tone of discourse within the open source community reached its nadir recently thanks to a tremendously unfortunate, and highly public, spitting match between Free Software Foundation leader Richard Stallman and Caldera president Ransom Love.
Neither man's comments have helped his own cause much. Still, I would have expected Love, a CEO and seasoned businessman, to have taken the high road when confronted with the insults of Stallman (who has never been known for diplomacy). It's clear that much is at stake--Microsoft is on the attack, Stallman continues to see Linux as an undesirable fork of the goals of free software, and Linux-based business plans are having a tough time succeeding. The passion is understandable, and I can excuse a bit of hyperbole and even desperation on the part of both folks. But calmer heads would advance Linux and free software far better than hot ones could.
Stallman's characterization of Love or Caldera as parasitic is no less dumb than fans of FSF-type licensing accusing Apple or Microsoft of stealing BSD-licensed code. You can't complain when you design a license a certain way and then groups you don't like use it as it was designed.
But Love's "what has Stallman done for Linux" retaliatory comments were no better advised. While Linux has strayed from his vision in many ways, Stallman helped form a foundation--in licensing and core software--without which Linux as we know it would not exist. I wouldn't go as far as those who would make Stallman a martyr while he's still alive, however.
If Love's comments on Stallman weren't enough to alienate diehard fans from Caldera, the company's new Linux products--and its associated licensing scheme--would certainly finish the job.
The new Caldera plan--which requires users to purchase a license (costing at least $59 each) for every desktop running OpenLinux 3.1--runs counter to the conventional wisdom that says that every Linux OS CD-ROM should able to be copied freely or loaded on multiple systems without extra licensing.
It's a risky move, but one that I neither hold in contempt nor deem doomed to failure. Caldera, for better or worse, is positioning itself as the company best able to mix the best of open source and proprietary code.
Rather than booing the company, the Linux community should cheer the fact that the Linux spectrum allows such diversity. At one end we have distributions such as Debian, which closely track current releases and don't consider any non-free code as part of the collection. At the other, Caldera not only allows but welcomes proprietary programs as integral parts of what it calls the OS.
While Caldera adds commercial components to its distribution, such as Borland's JBuilder Foundation, the company justifies its licensing scheme on more than just bolted-on components. The company is also talking up its "self-hosting" procedures for building its distribution, a technique that offers an enhanced level of integration to ensure that Caldera can avoid conflicts between different packages. It's more of a production technique than a feature, but it does provide some real value to those wanting a tightly-integrated distribution.
The company also promises greater attention to security; in all, Caldera promises a stable, self-contained OS kit whose packaging style resembles proprietary OSs more than other versions of Linux.
Of course the self-appointed Linux purity police don't like that. So what?
Caldera has never been the darling of the zealot set, but it's never really needed them because it's always targeted itself at the business market. Unlike any other Linux company, it has a large worldwide network of value-added resellers (VARs), many of which came from the company's acquisition of Unix vendor SCO. These VARs have until now been cool to Linux, or in some cases actively hostile. If they are to be expected to deploy Linux at customer sites, they must have their cut. And you can't give the channel much of a cut if the software can be freely copied.
So Caldera, with its licensing change, not only proposes a revenue model for itself, but one for its channel as well. It won't get existing Linux users to switch distributions, but it's not trying to. It can ignore the installed Linux base if it can attract current users of Windows or Unix systems. It has a chance of succeeding in businesses that have never before used Linux, who trust their VARs to provide a solid platform.
Caldera must prove that its new approach offers value and stability to corporate users trying Linux for the first time, and it needs to support its extensive reseller channel at least as well as SCO did. If the company can adequately tackle these two significant challenges, the insults won't matter. Caldera will have carved out a nice little piece of the Linux-in-business market.
Do you think Caldera's strategy has a chance? Tell Evan in the Talkback below or in the ZDNetLinux Forum. Or write to Evan directly at evan@starnix.com.