At the pinnacle of his rule the Emperor Napoleon had tens of millions of subjects and a domain which stretched across Europe to Russia’s doorstep.
But then Napoleon, a man who believed he was uniquely held in the hands of destiny, made a critical mistake. In defiance of the French continental blockade Portugal had been trading with his most hated enemy, the British, so with the consent of historical Portuguese rival, Spain, Napoleon invaded. But while he was there, thinking he could elevate the Spanish from what he saw as peasantry, squalor and superstition he also decided to conquer Spain.
Now, Spain was never the objective. Choking off trade with the British was. Spain was just an add on. Mission creep.
But what the invasion of Spain did was open Napoleon to years of bloodshed as guerrilla fighters waged an on going struggle against his troops. They’d move against him, he’d annihilate one of their cities but they just kept picking at him no matter how many of them his cannons smashed. Aided by the British, Spain repelled Napoleons forces. Napoleon in turn would have to drop what he was doing elsewhere and focus on the Spanish problem.
This bleeding war of attrition went on for seven years right up until Napoleon was sent into exile. As a man with a sweeping sense of his own greatness, in one of a few moments of absolute self awareness Napoleon pinpointed this peninsular war as the beginning of his ultimate downfall.
It was his great mistake.
If it hasn’t occurred to him already I’m thinking Larry Ellison is going to find himself thinking the exact same thing about buying Sun.
Sun wasn’t the objective. Being the largest most powerful software company on the planet was the objective, but then IBM came sniffing around Oracle’s incredibly large Solaris customer base when they started talking to Sun about an acquisition. The threat of IBM sales reps talking about replatforming that gold mine customer base from SPARC onto POWER must have loomed large. ‘And sure, we’ll throw in some DB2 licenses as we’re at it. Just to get you started.’ Having struggled and won the database wars this must have been a nightmare scenario, what was settled before could be going up for grabs once again.
With Mark Hurd not willing to take Sun’s hardware division off of Oracle’s hands (Knowing the true value was in replacing HP-UX with Solaris, and that wasn’t on offer.) Larry decided to go into the hardware business. A much, much, lower margin business than Oracle was used to (What do you think the reaction was when Oracle first saw STK tape margins?), but it came down from the top that they were going to make this work. They were going to invade the hardware market and elevate it from the peasantry, squalor and superstition software companies traditionally regard the ‘dumb hardware business’ to be.
But Sun’s customer base was already decaying. Oracle never had a hope in hell of clambering out of the muck and taking on the x86 server vendors, all of whom race to the bottom on deal pricing with looking to make up the razor thin margins in an avalanche of volume sales. SPARC was and remains stalled. Creating chips is like writing software if you’re a FABless company, but like software companies talent flocks to talent. Post acquisition they’ve been flocking away from Sun.
Plan B. Make up the money from what Oracle only wanted in the first place, the software. Solaris lost to Linux and is now something that can only be milked, a lot of the Oracle acquisitions are just maintenance cash cows so this fits fine with the business plan but it doesn’t give the company the pop it needed right away. How about a Java suit against Google Android? You’re then in the whitehot handset market without being in the handset market. Steve Jobs having chosen battery life over Java made sure it wasn’t in the iPhone so the only large new revenue stream would be from Android. A pay out there would go a long way to making Sun a balance sheet neutral transaction and start them into profit. Then there’s the reoccurring revenues to factor in. You could go after everyone shipping any form of Android with the Dalvik VM included. Now we’re talking real money!
Google win the suit and Oracle agree to a settlement of $0 so they can clear it and retry the case. But the Judge who made the ruling in the first place made it even more difficult to press the issue this time around as he writes a detailed legal judgment to hinder another such suit.
And then the hardware peasants start rebelling.
Now that Oracle has invaded the hardware market, their latest hardware BFF since the last hardware BFF has the gall to sue them for contract breach when Oracle stops supporting a processor architecture we all know is dead. It was in no way unreasonable for Oracle to stop supporting Itanium. It’s dead. It’s not coming back. Sales can only go down. But Oracle couldn’t even win that case because they had a deal in place they tried to hand wave away because it didn’t suit them to honour it.
With two black eyes from failed legal judgments resulting directly from decisions made as a result of the acquisition of Sun and a bloody nose from SAP quibbling the TomorrowNow lawsuit down to a reasonable number, suddenly Oracle is looking mortal. Mortal and without any allies.
These are just the opening battles in what will turn out to be a multi year war of attrition for Oracle. Like Napoleon, Oracle will crush some enemies decisively and will have to beat a dignified retreat against others but this is the great mistake playing out in front of us.
It was so much easier for them when they decided they wanted to beat Microsoft, now they have to beat everybody.
When they don’t beat everybody they will be overwhelmed and destroyed.