How Oracle Could Own Linux and Own the Market

Rob Enderle

Back when I was a competitive analyst, one of the projects that I enjoyed the most was writing scenarios. We would look into the future, define an event we either wanted to avoid or have happen, and then write how it came about in the past tense. In other words, let's say you were IBM in the 1980s (which is where I did this) and were concerned that Microsoft might become dominant (which few believed at the time). You would then write as if it were a few years later and Microsoft was dominant. How did it get there? You would then use this document to make sure that critical decisions that would enable this outcome were blocked and decisions that would prevent this outcome were encouraged. In our case at IBM, this went to the core of why my organization tried to spin out the IBM software division as a separate and independent company.


Unfortunately for those of us on the project, we were both right and the response failed to get the needed approvals because hardware believed they couldn't survive without us. This showcases that the practice can actually be very predictive but not necessarily effective at avoiding the predicted outcome. It is still a great deal of fun to do, though.


Oracle Dominance


Earlier this week, I wrote about what appeared to be the lowest risk path for Oracle with respect to its Sun acquisition. However, Oracle is not a low-risk company, and Larry Ellison, its CEO, is one of the most predatory CEOs in the world. This suggested another path, one where Oracle could end up in a position similar to where IBM was at its peak and with power that exceeds Microsoft's. So, taking this as a premise and moving to 2020, what would have had to have happened to get it there?


Keeping Hardware


First need: hardware control. Any path that doesn't include hardware doesn't get you to anything but some level of parity with Microsoft -- and you have to go through Microsoft to get there. Going through a large dominant company is nearly impossible. You have to find a way to go around, like Microsoft did to IBM. For that, Oracle would need to retain hardware and grow it to a level of dominance using tools unique to the new model. How it would get there is through hardware subsidies pulled from software revenue. Somewhat similar to what the carriers do with cell phones and the opposite of how IBM started, where software was 100 percent subsidized by the hardware, this idea of free hardware coupled with software annuity contracts could cut through a market starved for capital like a hot knife through butter. Given that Sun hardware is still perceived as premium hardware and Oracle is perceived as a high-quality vendor, this should work. One issue would be anti-trust problems once Oracle got to about 70 percent of the market, but this is likely the only path that might get it to 70 percent quickly. You'd want to pray for whoever was doing the logistics and I'll bet that if this breaks, this will be the cause. But this could be done.


Google or Apple Client: Could Ellison Be the New Steve Jobs?


Only two companies have viable non-Microsoft clients in the segment at the moment. Google, which uses a Linux-based platform called Android, and Apple, which uses a UNIX-based platform, the MacOS. Initially, I thought only Google could play but then it hit me that Larry Ellison is on Apple's board and Apple has a serious problem with respect to Steve Jobs' health and tenure at the moment. Larry and Steve are also reported to be friends. An Oracle-Apple partnership/merger, as a result, is actually much more likely than I initially thought and the combination has almost no overlap. Oracle could validate the MacOS for the enterprise, Apple could help Oracle with marketing, and Oracle would suddenly have a hardware portfolio in line with HP or Dell. Google, on the other hand, doesn't get it any hardware, and would be vastly more difficult to merge or partner with. In addition, there would be even more difficult culture and control issues involved in making this work. Finally, both Steve and Larry really want to take Microsoft out personally and their time is running out to do this. Granted, so does the Google leadership, suggesting the possibility of a three-way alliance, though I don't think that is necessary for this to be successful.


Owning Linux


If we go the Apple route, the MacOS has massively outperformed Linux on the desktop and the core aspects of both offerings are similar enough that Oracle shouldn't have to merge them. If it went the more difficult Google route, Android is a version of Linux and likely to become the dominant desktop distribution. But Unbreakable Linux, to give Oracle the control it needs, would have to either be the dominant or the only enterprise-level server Linux distribution. To make this outcome work, it needs to have a similar level of control to what Microsoft enjoys with its platforms. Granted, it would get there vastly differently, but Oracle has shown strong competence in this area. Control does not have to equal ownership but, with Oracle, I would expect that connection.


Wrapping Up: Oracle, the New IBM


If Oracle were to be as dominant as IBM was by 2020, this is the only reasonable path I can see that will get it there, and clearly this will be no walk in the park. However, it would be incredibly difficult for Microsoft alone to block, and cutting the deals needed to do so may be impossible with the current anti-trust restrictions on the company. While I'm not actually expecting this outcome, I did think it would be interesting to discuss. Hope you agree.

Subscribe to our Newsletters

Sign up now and get the best business technology insights direct to your inbox.


Add Comment      Leave a comment on this blog post
Apr 24, 2009 1:32 PM obvio capitao obvio capitao  says:


Apr 24, 2009 2:07 PM Rob Enderle Rob Enderle  says: in response to obvio capitao

Funny story, and I'm actually not kidding.  When we presented the scenario on Microsoft eclipsing IBM in software 6 years before they actually did that is the exact word one of the top IBM VPs used.  He went on to say that there simply was no way a company that made software for toys could even sell into the enterprise without IBM let alone pass us.  

Just so I have this for historical reference, what part of this did you feel was poorly founded?   Don't get me wrong, as I wrote, I think this outcome unlikely but just in case something like this does happen this would make for a great footnote in a book.  

Apr 25, 2009 10:12 AM obvio capitao obvio capitao  says: in response to Rob Enderle

Even a stopped clock is right twice a day.However your predictions look much more like a clock that is always late:In 2003 you said that "the biggest long-term problem with moving to an Apple platform is that the company is in decline". (Apple stocks rose almost 1500% since then.)After that, you said that Apple would not benefit from iPhone ("too expensive in a highly competitive and saturated market").And, more recently, the entire "SCO owns Unix" fiasco.You are so clueless and frequently wrong that Wired once concluded an article about "The 15 Dumbest Apple Predictions Of All Time" with this pearl: "Any others dumb predictions to add? Now, don't go posting the complete works of Rob Enderle, folks. Bandwidth isn't free."

Apr 25, 2009 4:03 PM Rob Enderle Rob Enderle  says: in response to obvio capitao

What is sad about attacks like this is that they prevent learning.   You are one of those people who clearly don't like to see opinions you disagree with and are so blinded by your anger you can't see straight.

For instance this isn't a prediction, and neither were two of the three things you listed.   In this instance I clearly write I don't expect this outcome.   In the case of Apple in decline that was in past tense responding to a reporter's question on why people weren't moving to Apple in business.   In the SCO example, taken out of context and reworded, that isn't a prediction either.   Even the time line you imply is disingenuous, all of this comes from around 6 years ago. 

If you have to misrepresent things, take things out of context, and create false impressions to make a point doesn't that suggest the point you are trying to make is false?   Why work so hard to fabricate things, why not try to argue on merit?

Even the link you used doesn't work.   If you are going to be critical maybe being right more than once out of 4 times yourself would set a better example. 

Was the problem you are apparently trying to cover up with this smoke screen and phony personal attack that there wasn't anything here you could really disagree with?   Are you truly that shallow? 

Apr 26, 2009 7:14 PM paul gleason paul gleason  says: in response to Rob Enderle


I don't understand your scenario for Oracle domination.  In particular, you seem to be saying that Oracle could and should take control of Linux.  That doesn't seem possible because of the IP licensing.  Anyway, they already have control of their own Linux distribution.  By making contributions to the kernel they are pushing Linux in the direction they want already.  In fact, Oracle is one of the top contributers of code to the Linux kernel.

It never ceases to amaze my how a few people on both sides, proprietary and FOSS, try to play fast and loose with the other guys IP.  But so far Oracle is playing by the rules, taking what they want from the FOSS world (RHEL, Apache), fulfilling the licensing requirements, and even giving back significantly to the community (btrfs, ext4 patches, etc). 

I think a far more interesting scenario would be for Oracle to continue as it has as a mainly proprietary company, taking from FOSS when it makes business sense.  This strategy alone could provide a significant cost advantage to those who don't follow it.  Of course you can count on HP and IBM (and a post Balmer Microsoft) following a similar strategy, so it wouldn't be easy.

Apr 27, 2009 8:28 AM Rob Enderle Rob Enderle  says: in response to paul gleason

What appears to be the core of your argument is that you don't want any company to control Linux so you believe it to be impossible.   But anything is possible.   The Chinese Government, in China, could easily specify and control Red Flag Linux.    Oracle can control Unbreakable Linux and, were it the last enterprise Linux distribution standing, by default, they would control (effectively own) Linux in the enterprise.   Are you telling me Oracle couldn't take out Red Hat and Novell if they put their minds to it?   Recall that PeopleSoft was much more powerful and it, in hindsight, wasn't much of a problem.   Granted another Linux distribution could pop up but, in this market, who would fund it?

They will shortly own the most powerful (doesn't mean as much as it once did does it) UNIX variant with Solaris and that gives them a foundation to build on.   Oracle will play by the rules, but they are far from an Open Source poster child.   They are probably the most aggressive company in technology at the moment and they play hardball to win. 

This is a land grab; the only question is whose land are they grabbing? 


Post a comment





(Maximum characters: 1200). You have 1200 characters left.




Subscribe Daily Edge Newsletters

Sign up now and get the best business technology insights direct to your inbox.

Subscribe Daily Edge Newsletters

Sign up now and get the best business technology insights direct to your inbox.