Oracle–Linux Knight that isn’t Quite

After persistent media rumours of an Oracle-based GNU/Linux distribution, Mr. Ellison finally announced it. Sort of. It’s offering Oracle support services around the Red Hat Linux distribution. It makes sense–I think companies need to make the entire life of the software solutions they sell a seamless continuum lacking problems and time-wasting intervention from their customers. Yet, a lot of people are writing about how this is a move to hijack Red Hat’s support business or is a forking manoeuver. Some are saying it’s a warning to Red Hat, which is true, but I think the warning does not come in the sense of a support business hijack or fork. Oracle must have a more comprehensive picture in mind. Rather, I think Oracle’s move makes sense in order to steer its solutions into a comprehensive and lean offering. So I’ll explain why the fork doesn’t matter and why the support services on their own aren’t the real threat.

I refer to an idea I put forward some time ago for what needs to happen with the Linux desktop to catapult it to widespread adoption. I was asserting that there needs to be a Linux vendor changing the entire OS market game by offering: A full computing solution should come from a company that pre-bundles everything its customers want, consistently supporting it, for the duration of ownership. It should not require anxious intervention from the owner when the owner desires a new component or new system, and the new system should have all data and applications from the old system installed, setup, and accessible upon delivery. I would expect this to apply just as well to business software systems as individual user systems. I further believe that such a solution is only possible to fulfill, in all its complete glory, within a FOSS ecosystem (read that idea link at the beginning of this paragraph for details on why). This latter point can play to Red Hat’s favour. How does this relate to Oracle’s Linux move? Well, let me clear the threats that have been proposed already, which I mentioned were passing around the blogosphere/news article space.

First, if Oracle intends to eventually fork Red Hat’s distribution, so what? It’s been done more than once before. Mandriva is such an example. The GPL licensing mechanism has proven over and over that FOSS forks do not necessitate a negative outcome. Often they lead to a spread of improvements in the ecosystem as a whole and all the companies that continue to participate in that spread tend to benefit. I mean, Red Hat is still around, and strong. Forking in the open source world ought to have earned a default view as growth rather than as problematic division. The freedom of the licensing schemes makes a FOSS fork wholly different than one in which the prongs are proprietary.

Second, I don’t see why this should be viewed as Oracle just trying to hijack just Red Hat’s support business. Maybe I’m missing something but that’s only one aspect of what is at play. The fact is that Oracle offers products that run on Linux. Red Hat offers products that run on Linux. Some of these compete (application servers, database servers, etc.). Furthermore, competitor Microsoft offers enterprise solutions and the OS. These companies are getting their tentacles around more comprehensive offerings for their customers. At least that’s what the advertising is always promising. So what is going to be the easiest, most painfree choice for a customer? Getting pieces of a solution from a combination of vendors? Or getting something from start to finish that eliminates

• further evaluation time and effort
• additional sales points-of-contact
• extraneous support sources when trying to solve problems
• integration woes

It seems to me that all of these eliminations would result in a much more attractive choice for the customer. It would save a lot of time and cut down on the efforts required to get well-designed and supported enterprise systems. Consider how Oracle has positioned this move, reiterating its “unbreakable” theme and stating that

“Oracle validated configurations provide easier, faster, and lower-cost deployment of Linux solutions in the enterprise with pre-tested, validated architectures–including software, hardware, storage, and network components–along with documented best practices.”

If this support move is Oracle’s warning shot, then I think it’s a shot of the lean and comprehensive solution nature. Their phrase, which I quoted above, is spot-on with the model I outlined for a real year of the Linux desktop, it’s just focused on the enterprise software space instead.

Considering Oracle, Microsoft, and Red Hat, who can actually do this? The all-proprietary vendor, Microsoft? The mixed proprietary/open source vendor, Oracle? Or the all open source vendor, Red Hat? Based on those three options, only Red Hat is actually in the position to fulfill this comprehensive model because it’s the only one that operates entirely in the FOSS ecosystem and for that, again I refer to the comprehensive and lean model, which I outlined previously.

I’ll leave this post with a question. Where is Canonical in this? Every blog and article talking about Oracle/Red Hat seems to ask or fantasize about Canonical and Oracle teaming up. Thus far Canonical is publicly touting only support services. Will it catch up on the OS market game change to offer the sort of solution Red Hat currently has the potential to? That Oracle seems to want to do? That Microsoft tries to do but limits itself under pounds of proprietary rope?