Skip to main content

What Larry Said: "I'll oracle you!"

Posted by editor on June 3, 2009 at 11:32 PM PDT

I haven't really said much about Oracle in my Editor's blogs. There are several good reasons for that, among them the fact that I know nothing more than what's out there publicly already; also, I don't consider fanciful speculation to be a fundamental element in my job role as editor. However, when Larry Ellison came to the stage as the climactic speaker in this year's JavaOne opening keynote session, it became incumbent among my duties as editor to report this event. So, this is my first and last blog on the topic of Oracle and Sun, until something is completed and I'm instructed to write more about this topic by those who pay me to do this...

It's of no benefit for me to quote what Larry said, since the video is readily available online; so, the only added value I can possibly contribute is my own view on what he said, and on why he was chosen (or chose?) to speak in that particular time slot. So, that's what I'll do in this post.

A bit of personal background...

First, some personal background, if you're interested. If you're not interested in my background, skip to the next heading. But, IMO, with something like this, where there's chatter expressing everything that can possibly be stated, you need to know who's talking before you can properly interpret the opinions expressed. If you agree with that, read on!

I've been a software developer going on 30 years now. Primarily, I've worked on the back-end, less visible side of computing -- lots of mathematical modeling, intensive analysis of high data volumes, server-side database development... Basically, my programming career has involved trying to get the maximum throughput possible, either to process, reduce, and categorize very large volumes of data, or serve large numbers of simultaneous users.

In this job role, I've worked with many technologies, including HP's RMBasic (an early scientific programming platform, designed for interacting with instruments through serial and GPIB interfaces); Sun multiprocessor workstations (starting in 1993, when doing "symmetric multiprocessing" put you into a software engineering vanguard); Microsoft server-side platforms, including SQL Server (near the dawn of the .NET age); and, yup, Oracle. The past 15 years of my software engineering experience have been dominated by high data volume, computationally intensive programming first on SunOS, then Solaris, platforms; and in the past 7 years an Oracle database has been the hub of that development.

So, in terms of the technologies and companies we're talking about, and which Larry was talking about, I do have a bit of experience. In addition (I'll say this, since some of you may already know it, and in the interests of full disclosure), I was the editor of BEA's developer sites last year, when Oracle acquired BEA and integrated the BEA sites into the Oracle Technology Network. This means that I personally know (in a business relationship) some people at Oracle. That, in fact, is another reason why I haven't said much about Oracle and Sun, and the future of, Java, and all the other topics that are receiving lots of speculative commentary out there...

My view of what Larry said

OK, so I really mean this disclaimer: everything stated here is my own opinion. I have no actual knowledge of what will happen. I'm stating all of this based on 15+ years of working with Sun technologies and 7 years of working with Oracle technologies. When you work with a particular vendor's technologies over a long period of time, you get a feel for what they're about -- you develop an opinion about that, anyway. So, here's my opinion about why Oracle chose to buy Sun, why Sun is pleased to be acquired by Oracle, and what it might mean for the future of, Java, et al.

If I was Oracle, I would see myself as the rock-solid database infrastructure in the world. (Well, IBM is right there too, but I consider IBM to offer a one-company solution, which means it's overall a less "open" platform than Oracle from a developer's standpoint.) As Oracle, I would see that I was the best rock-solid database, and know people and industries need what I can provide. But I would also see a gap: everything I do is rock-solid, the quality, the reliability, the efficiency, are stellar -- yet, there is a bigger world out there! Not everything happens in the Data Center.

"What if," I'd think (if I was Oracle): "what if I could also bring the same reliability and performance that I have in the Data Center out to the smallest devices? Because, isn't the world of devices that connect to Data Centers getting smaller and smaller? Aren't 'computers' disappearing into devices that are known not as computers but by names like 'smartphones'? Aren't these the consumers of my Data Center's activity, as well as the providers of the input to my Data Center cloud of the future? What if I could have a presence there too?"

So, what did Larry talk about? I heard talk primarily about software on small devices. Small devices are an area where Oracle has little presence today. But, think about Oracle's primary competitors, say, IBM, or Microsoft. Don't they have a much greater presence from the desktop on down to smaller devices, where Oracle is largely absent?

Larry has been around for some decades, and has seen how the fortunes of even major companies can change in ways few would expect to see. In the late 1990s, I used to advise people (at and my now-silent site) that the stocks of companies like Sun and EMC were likely much overpriced. I thought that was the case because it seemed unlikely to me that the value of companies could quadruple in two years (which the Nasdaq approximately did in 1999 and 2000) while the value of their primary customers (whose stocks were on the NYSE) were rising at an incredibly more subdued pace. There was a mismatch, an economic/financial discontinuity. Either the Nasdaq stars were highly overvalued, or their NYSE customers were incredibly undervalued...

So, Larry watched this, navigated Oracle through it. He also has seen the rise of Linux on the desktop. I'm writing this blog on an HP Mini 1120NR, a Linux netbook. A few years ago, who'd have thought that could come to be? And no, I didn't install Linux on my HP Mini: it came that way.

A big news item at this year's JavaOne was the new Java Store. Java is installed on an enormous portion of the world's computers. Oracle is a "Java shop," but of course they need to be much more diverse: they need to provide a reliable, solid database to all developers, not just Java developers. Similarly, Solaris is a superb platform for Java and also for numerically intensive programming in Fortran, C, and C++ (all of which I've done lots of).

To have Sun under the Oracle umbrella, IMO, provides Oracle with a more direct connection to platforms where it previously had a minute presence. In particular, the small devices. But, also, with the development of OpenSolaris (I recorded a podcast with that team today), a potentially significant future presence on the desktop.

For Sun: being under the Oracle umbrella provides a "permanent" future. I mean, we've seen great companies go out of business, or become dependent on government bailouts recently, no? I don't think the U.S. government would consider Sun Microsystems worthy of a bailout, if the company came to desperate straits.

Again, for Oracle: an IBM/Sun company would have virtually closed the door on Java to Oracle. An Oracle/Sun company presents some issues for IBM, IMO. Just as Oracle/Sun might induce some thinking inside the suites of Microsoft. And, realistically: isn't a well-funded OpenSolaris a challenge to Linux?

So, strategically -- when I heard this news (I heard it first at an O'Reilly weekly team meeting), my first thoughts were: "Wow, Larry really is a genius!"

But what about the communities?

The big fear I am hearing, seeing in my browsing of the commentary, is that Oracle is too different, they simply don't understand the community aspect, they're all a bunch of corporate types selling software for big bucks, they don't understand open source, they have no interest in openness, they're going to crunch JavaFX out of existence, Java is going to turned into Oracle's pet, will be absorbed into the Oracle Technology Network because Oracle only understands the "resistance is futile" mode of operation, etc., etc., etc.

I disagree with all of this. Why? Because I think Larry, and Sun's management, view the situation from a much broader, long-term point of view. IMO, Larry (and his team, it's not all just Larry, of course) have always had a great sense of where technology stands today, and what's lacking today that will be needed relatively soon. Furthermore, Oracle has demonstrated the ability to act upon that vision, and be there when the rest of the world realizes it needs that something, saying "Oh, you need this now? Here! Here it is!"

Jeff Bezos (Amazon) is another person who I think has the trait of foreseeing future needs with precise clarity (e.g., development in the cloud). People like Larry Ellison and Jeff Bezos invent things that we regular people consider "interesting experiments" at the time -- then five years later, entire technology conferences are focused on how everyone needs to get on board utilizing those services ASAP--else, your company won't survive.

I'm still not effectively addressing the communities issue that I know is so much on people's minds. I'll try to do that now (this is getting a bit long, I apologize for that -- as I said, I won't be repeating discussion of this topic any time soon).

IMO, Larry has watched the development of "Web 2.0" and has watched computers on the client side "disappearing" into devices that are not considered to be "computers." I argue that someone like Larry doesn't want to sit solely in his Data Center kingdom watching a major technological wave proceed in the opposite direction. Of course, that Web 2.0 wave can't get anywhere without the Data Center. But, still, it's got to be enticing to actually be able to participate in it, ride that wave, race in it along with other competitors. Who's going to ride that wave best? Will it be Microsoft? Apple? Linux? IBM? But, wait a minute... Hmm... "Why not Oracle?" Larry muses...

"But Larry has no experience with this kind of wave," you argue. "He knows nothing about communities, about open source, etc." Well, I remember a time, a decade or so ago, when Oracle and Sun were united in what I then considered screaming a bit too much over Microsoft's practices and dominance of certain areas of technology. My point is: there has been an understanding and respect of point of view between Oracle and Sun for a long time. My guess, my prediction, is that Oracle is highly unlikely to ignore something Sun tells them is important. My estimate is that Oracle is doing a lot of listening right now.

At JavaOne, I learned that Sun has laid the groundwork for a long-term future for Java, by "seeding" the universities with Java education. Education is a major component of the activities that occured this week at the booth. My son did his honors program software development at Wesleyan University in computational microbiology in Java. Universities are on board with Java, and the upcoming development community is also accustomed to open source software and tools.

Meanwhile, these university CS students know nothing of Oracle. They may know MySQL or Postgres. Once they graduate, they'll enter a world where greater robustness and reliability are required in the technologies that are the center of their daily endeavors. If a Data Center is involved in their work, they will likely come to know about Oracle. Conceivably, they will some day be decision-makers when the question of "Linux or OpenSolaris?" is brought up. But this question won't arise without a viable, well-funded OpenSolaris effort. As we know, Linux is quite well-funded today, even though early on it was indeed one person's creation. OpenSolaris will need strong backing to be able to compete.

"But--" (you keep saying this!) "Oracle doesn't understand communities!" Well, how about this response: the people who create the Oracle database don't need to change their already successful mode of operation. Oracle can still sell their database the way they always have. Java's independence doesn't challenge that by any means. All that's necessary in order for the worries of the Java community to be allayed, is for Oracle to say to Sun "I see what you mean. That is quite valuable. Yes, you keep on doing that community stuff. And, by the way, we'll increase your funding for that effort as well." And I think there's a really good chance that that is indeed what's going to happen. And over time, parts of Oracle may even become more like parts of Sun, moreso that vice-versa. That's my prediction. Oracle will learn stuff it doesn't know about from Sun. And Larry wants that to happen, IMO.

"I'll oracle you!"

Have you "googled" anything lately? Grabbed a "Kleenex"? I think Larry would be quite pleased if someday, as you and your friend part, one of you says "I'll oracle you!" -- meaning they'll get back in touch with you through "oracling." I have no idea what "oracling" someone in the future might entail. But I'll guess that it would involve Java technology on a device possibly manufactured by Sun (but also competing devices manufactured by others) that connects to Data Centers that might operate on Solaris servers running Oracle databases (but also runnable on other platforms)...

To me (in closing), a fusion of Sun and Oracle opens up the possibility that some day (maybe 5-10 years from now) we will indeed "oracle" one another. Without that fusion, I don't think that possibility exists. Oracle apart from Sun gives us Oracle-powered this and that, but we'd never "oracle" one another. Oracle/Sun creates a rock-solid technology continuum the spans the biggest data center down to the smallest device, and reaches out to the people who use those devices, and involves all the communities that are involved in bringing those services, the software, the gadgets, to the on-the-street users.

Call me naive, unaware, or uninformed -- whatever you'd like -- but that's the way I see it.