Monday, April 20, 2009

Oracle Comes To Rescue, Buys Sun

According to a joint press release, Oracle has saved Sun. Well, the press release doesn't say much about saving, but does outline the acquisition of Sun by Oracle.

Same thing really. From an outsiders perspective, Sun was such a horribly run company since the 1.0 bubble burst, it could not survive without being saved.

The truth is Sparc was a great thing in the 90's, before Intel really got going. But since then Intel, initially spurred by AMD, has totally blown the rest of the chip market, such as it was, away. This just makes sense. They are the only company that had the resources to invest in fabs (factories) at the necessary scale. No one else could possibly keep up.

This left Sun with this massive customer base and investment in a technology that had absolutely no future. They tried to shift towards Intel chips, but you could tell their heart wasn't in it. Then their once vaunted Solaris operating system got whipped by Linux. Yes there are some things Solaris can do that Linux can't. But Linux was open source, which means it just had more team members. Solaris went open source, but by then it was too late. It was another also ran along with Sparc.

Towards the end, Sun move aggressively to position around software, and specifically Java, even, oddly, changing their stock ticker symbol to JAVA. This was strange since, for one thing, the company makes so little money from Java, and they never executed or even articulated a strategy to do so.

Along they way they bought MySQL for a billion dollars. More folly for a company whose just announced sale price was 7.4 billion dollars to Oracle.

Perhaps it is unfair to blame Sun's failure on bad management. They had great success in the 90's, but the anchors of that success in the 90s - Sparc and Solaris, became anchors around their neck after the bubble burst. Sun is a big company and it is indeed hard to shift away so quickly from all the things that were so recently the core's of your success.

In truth, in thinking about this, I am not exactly sure what Sun management could have done given the hand they were dealt. What do you think?


  1. I'm more interested to hear your take on MYSQL now belonging to Oracle, any thoughts Hank?

  2. I don't think Oracle's ownership of MYSQL will have a negative effect, and it might have a positive one.

    The negative view is that Oracle has just taken out a competitor. But I dont think even oracle can put the cat back in the bag. There is still Postgres, and all the MySQL source code is public. And I think if they tried to change stuff, more developers would leave.

    It will be interesting to see whether Oracle can find a way to Make MySQL make more money without hurting their core database franchise.

    My view is that relational databases have become a commodity that is only protected by lock in. There is little innovation in the field anymore. Only the super big companies that have products that are dependent on specific oracle APIs will continue to pay. So I think Oracle, broadly speaking, has a long term problem on its hands. It is addressing this by moving up the stack to applications where they can add more unique value.

  3. I'm surprised Sun lasted as long as they did. I started to count them out in 2004. Unfortunately my co. is still deploying solutions with SPARC/Solaris systems

  4. Sun started failing as long ago as the messed up Unix International fiasco that motivated all their competitors to join against them and AT&T. I think that was something like 1990. If only they hadn't been arrogant idiots at that point, they could have maintained much more market share during the mid and later 90s. Furthermore, that fiasco prompted the emergence of Motif as a UI standard, which was possibly the worst-ever windowing UI toolset in terms of both ease of development and usability.

    The SPARC architecture was never more than mediocre. I mean, sure, it was better designed than early Pentiums, but really Sun coasted on servers for a long time due to the utter inadequacy of '90s Windows platforms to run servers efficiently. But in the end, Pentiums scaled much better than SPARCs. Intel could afford a lot more fab plant capital for millions and millions of chips than Sun and partners could for I guess tens or hundreds of thousands.

    Anyhow, 90s-era SunOS was a nice development platform, and they had even more dominance and superiority in the late 80s when there were no other sensible alternatives at all.

  5. well well i wanna buy a soccer ball

  6. Why not cooperate to achieve a common goal. Anyway, the main purpose of their existence is profit related but with the effect of giving good customer service to remain in the market no matter the situation.

  7. I use Linux and Solaris on a regular basis. I really like Solaris on x86/64 from a systems admin/management perspective, but you can't get away from the fact that a lot of software is linux only on linux biased. By that I mean even downloading source and compiling may require tweaking files to get the software to build, much the less run as fast as it should if it were correctly optimized and tuned.

    My thinking is this will be like the acquisition of DEC by Compaq and then by HP. Whatever jewels are left in the Solaris crown won't get the polish they deserve until they're not really worth all that much, anyway. I think you're right that SPARC is a dead-weight, but I think that Solaris is good stuff. With 20-20 hindsight the time to start giving away Solaris on Intel was when Linux started making headway in the late 90's. Of course, everyone was getting fat on SPARC at the time...

  8. Get your teeth in my son... Well executed slashing. If only we stood back more often we'd see more occurences that suck!

  9. Its so sad that Sun is sold!! :( I loved the business model of SUN. It was all open source and it was great.. we just need to hope for the best for sun..

  10. thansk for artickle
    i think is this will be like acquiston dec by compaq and then by hp.

  11. Sun committed suicide with this open source and Java nonsense. They deserve to die - good riddance.

  12. I am just hoping that Oracle keeps maintaining OpenOffice. This is a strategically important piece of software.


Note: Only a member of this blog may post a comment.