Unbreakable, broken
So Oracle, as far as I can tell, has got it’s hate on for open source companies. It is going for the gusto with Unbreakable Linux and potentially with Unbreakable MySQL. It already fired a shot directly into MySQL’s bow with it’s purchases of innoDB and sleepycat - two key elements of MySQL’s landscape.
I want to look at the Unbreakable strategy. The idea is that Red Hat (and MySQL) hand out all their hard work for free, so Oracle can simply repackage that work and sell it with support at a fraction of what Red Hat would charge. On the surface this seems like a killing blow to the poor free software vendors, Wall Street certainly thought it was and cut around 50% of Red Hat’s stock price.
So let’s look at how Unbreakable is going to work. If I am a company and far from looking for a free solution (amongst the innumerable free distributions) I am looking to buy support for a linux distribution, what I want is to have updates come as quickly as possible and if I find a problem I want the vendor to be able to address my problems. Now with Oracle there is a built in time delay for Oracle to roll out new Red Hat changes. More importantly there’s several issues with Oracle’s ability to go into the code and fix things.
If they modify the source, they then have to also wait for Red Hat to fix it and then remove their fix and insert Red Hat’s. If Red Hat in the meantime rolls out changes that do not include Oracle’s fix, they’ll have to actually add the fix to the new version as well and then refix when Red Hat addresses the problem. The customer may have to update for each of these iterations.
On the other hand if Oracle does not wait for Red Hat changes and simply starts updating Unbreakable with their fixes and changes then, all of a sudden, they aren’t repackaging Red Hat anymore. They have their very own distribution for which they have to do all the work that they were going to have Red Hat do for them. Then they have to staff up to handle that and now their super low price becomes either a continuing loss for them or they increase their price at which point they are competing more directly with Red Hat, with the downside for them - of course - that they aren’t actually in the Linux business.
On the other, other hand.. let’s look at this longer term. Let’s say pigs start flying and Unbreakable becomes a raging success. It turns out that every company is actually looking for bargain basement support for potentially key infrastructure elements and Oracle eats Red Hat’s lunch in it’s entirety. That puts Red Hat, you know, out of business. Then who does all the work for Oracle? I suppose Oracle could start doing that work, but now all it’s customers are going to be saddled with the same price they were with Red Hat from a company who’s only goal was to crush its competitor. So they will end up running a distribution that used to be developed by a company for which it was the entire focus of that business, to one that is developed by a company who wanted to undercut the original company and for who Linux is a tiny portion of it’s income. Honestly, I don’t see where the long term proposition is in Unbreakable. By it’s very existence it is antithetical to the product it is trying to sell. At best it is a non-event and no one uses it, at worst consumers and customers lose and really only Oracle wins.
I think everyone is beginning to realize this problem with Oracle’s situation - no one is switching, the few who did did so because Oracle offered an unpublished additional 50% off their already low price. And understand that at that low price, even Unbreakable becomes a loss for Oracle.
I think ultimately Unbreakable is actually a boon to the Free Software movement by proclaiming very loudly that Ellison is worried about these companies and is taking moves to try cut competition off at the knees. This, in it’s backhanded way, adds legitimacy (if indeed any more was actually needed) to Red Hat and perhaps MySQL. That, at least, is my hope - because the success of Unbreakable would be a terrible thing.
