So what can Sun do?

So what can Sun do?

Summary: Sun is the Rodney Dangerfield of the tech companies on the stock market - no matter what they announce, from sales gains and new products to losses and layoffs, the response is always the same: the people who make money selling IBM and Wintel shares short Sun. It's a loser: the dollars at issue favor the MS/IBM side by over a hundred to one - so it's time, I think, to pick up that ball and go home. Take the company private and be done with this nonsense.

I don't believe Sun can rescue its share price by winning the media war - the failure of its earlier three part strategy: share buybacks, a reverse split, and the JAVA acronym change - illustrates that. The "reality disconnect" here is nearly total: if the market followed value or earnings potential, that strategy should have combined with new products and new markets to produce a rising share price. It didn't - and, whatever the reason, that's the bottom line: it hasn't worked.

So now what?

Continuing to do what doesn't work is traditional, but absurd - so it's time to reach for sensible alternatives like taking the company private. One option? a lightly leveraged employee buyout taking advantage of both the available cash and the company's low share price.

And after that? My guess is that McNealy's single arrow strategy would make the most sense.

That means divesting the x86 business without giving up the customer base - and, on the positive side, a new public company spin-off can both facilitate the buyout and provide the legal lubricants needed for significant organizational (i.e. staffing) change.

In the alternative, there's certainly room to talk about off-loading x86 retailing and support to any of the big three: HP still has some die hard Unix customers it doesn't want to support; Dell is increasingly desperate for a market differentiator against both HP and Acer; and IBM has a Lenova problem.

The new company would put all of its wood behind the coolthreads products and their successors - and I do mean all of the wood. It's increasingly obvious that open source is on a run away adoption curve and will ultimately relegate companies relying on software licensing revenues to history's scrap heap. Sun is well along that road now, and could jump further ahead: develop a strategy based on selling only hardware and operational support with no separate consulting, no separate software licensing, and no mixed messages for customers.

Basically: the strategy would be to charge a premium price for a premium product and differentiate on operational stability.

Build on the Fujitsu relationship, stay the course with the big data center customers, but focus a lot more effort in the small to mid range market: the people who desperately want stuff that "just works" and now think they have nowhere to go.

These people are angry about IT, about something that looks so simple but costs them money and aggravation every single day of the week -and no one's telling them that Sun has exactly what they need. Bundle open source applications with preconfigured hardware, provide personalized sales and real operational support, add legal protections, push Sun Ray, and these guys will stampede to Sun.

And that, particularly for the people who work at Sun, is really where the silver lining is for today's market behavior.

Topics: Oracle, Banking, CXO, Data Centers, Hardware, IBM, Storage, IT Employment

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • I have a couple of suggestions.

    Now doubt you wont like them, Murph, but here they are anyway:

    - Work with Microsoft to get Windows Server running on SPARC. Why? There are a lot of small businesses on Windows servers for whom the migration to Sun hardware is simply too big. Being able to run Windows has worked well for Apples fortunes in the last couple of years.

    - Put a rocket up Gosling's arse and pick up the pace on Java's evolution. It's as good as static right now, and there's plenty on interesting stuff going on in the languages space at the moment.
    • first point

      windows NT used to run on intel but also alpha, power pc mips(beta i think i remember). that was an utter failure, and I doubt microsoft is willing to try it again.

      Neither would sun as they couldn't exist on their sole hardware revenues.
      • Windows on Intel, Alpha, MIPS

        Yep, Windows NT was multiplatform. Was shown on PowerPC, but IBM/Motorola couldn't face the reality that their options (Mac pre-OSX and OS/2) didn't have the right stuff (true muti-tasking) so they didn't didn't shell out for the HAL development.

        McNealy demanded Office for Sparc, and Gates/Ballmer laughed in his face. When AutoDesk dropped SPARC, the workstations went Intel, and I thought that Sun's fate was sealed.

        That they have survived this long is a minor miracle, since the other Unixes (except OSX) are dead. Perhaps if Sun had delivered on one promise (Solaris on x86 when Windows was vulnerable, "write once, run anywhere", etc.) instead of relying on litigation (didn't Microsoft have a license and the best x86 JVM after all) they might be relevant.

        Also-ran, like Murphy.
    • It doesn't make sense to port Windows to SPARC.

      Sun appears to be abandoning SPARC on the low end as it is uncompetitive in single thread performance as well as price compared with x86 systems.
    • customers are not doing that...virtualization!!

      it goes something like this...

      windows server - running on
      VMware/virtual os something or other - running on linux

      running on hardware platform that is fast and cheap....

      IMHO the days of running windows servers directly on hardware is dwindling in large to medium enterprises.

      • virtualization does not abstract processor architecture. nt.

        • You can do this with Qemu.

          Qemu will do this, though you will need to check if the
          processor you are targetting or needing has been

          "QEMU is a generic and open source machine emulator and

          When used as a machine emulator, QEMU can run OSes
          and programs made for one machine (e.g. an ARM board)
          on a different machine (e.g. your own PC). By using
          dynamic translation, it achieves very good performances.

          When used as a virtualizer, QEMU achieves near native
          performances by executing the guest code directly on the
          host CPU. A host driver called the QEMU accelerator (also
          known as KQEMU) is needed in this case. The virtualizer
          mode requires that both the host and guest machine use
          x86 compatible processors."
      • Virtualization Indeed!

        Non-Windows server manufacturers have an issue -- virtualization vendors are eliminating them. Is SPARC the best option for VMWare? Not that I recall.
    • Java has been open-sourced...

      ... or, more specifically, donated to IBM with a few licensing restrictions. Which company, IBM or Sun, would be better able to make profitable use of Java improvements?

      Add to that the significant adoption of .net as a restriction on market growth, and there's no potential advantage for Sun to make a major investment in Java.

      Open-sourcing is a means of continuing development while reducing staffing expenses. To make a major investment after open-sourcing would be contradictory.
      Anton Philidor
      • Observations based on a misunderstanding

        Whilst Java continues to develop, jump over to, the big areas in Java
        development are higher up the chain.

        Check out the Eclipse or Netbeans, various Java application
        Servers (JEE5), Java projects at Apache, etc.

        Opening the source to the JDK is enabling Java to be
        ported to more architectures without Sun or partners
        having to carry most of the load. This frees their
        employees to add significantly more value with new ideas
        and contributions.

        Java development is today healthier than it ever has been.
        Richard Flude
        • You haven't disagreed.

          My point was that Sun would gain little from a major investment in Java because the company had sacrificed its ability to make a profit on the software. In order to improve the share price by laying off the people with new ideas and contributions to make.

          That IBM, which had been taunting Sun for years about open-sourcing Java in order to save the licensing fees, has used the free software to create products like Eclipse doesn't matter to Sun. IBM can use Java to reduce payments to other software companies as much as it chooses to do so. Without paying Sun.

          Java will continue to be in widespread use for some time. In much like its current form, I suggest.
          Anton Philidor
    • WABI - and Yes, sort of

      IS MS wanted to put their apps on SPARC the easiest way to do that would be to re-enable WABI. That would mean licensing change and some applications change but would otherwise be fairly simple and yeild the ability to run MS apps on SPARC without recompilation or code change.

      Back when Windows 3.0 first cam eout and the PC was struggling to get to 600 x 800 screens WABI, on HyperSPARC, ran stuff like Excel much faster than the PCs did - at 1280 x 1600.

      Second: I agree JAVA is stagnant, but I think that's because it's pretty much dead ended: further change will just make it more complex, not better.
      • Java complexity

        You're right about Java's accretion of complexioty in the past. But couldn't it be improved by rationalization and simplification?

        If there were a profit in it, wouldn't a more efficient, easier-to-use Java be an advantage?

        You wrote:

        I agree JAVA is stagnant, but I think that's because it's pretty much dead ended: further change will just make it more complex, not better.

        [End quote]

        A product which couldn't be made better? That would be a first, no?!
        Anton Philidor
      • Why would MS push this?

        I think Sun would gain more than Microsoft.

        Rather than bugger about with WABI or porting Windows, Microsoft could surprise us all provide a .Net runtime for Solaris. I don't see it happening anytime soon, though.

        Where Java has stagnated, .Net/C# has not - it has continued to improve without unwarranted additional complexity.
        • I don't think they would

          But you suggested Windows on SPARC - I only said that WABI would be a better route to the same ends.

          In fact I keep wondering about this because the Sun/MS mutual licensing agreement that came out of Sun's java lawsuit should enable Sun to force this - and that would be really cool.
          • wabi is win 16

            and wabi has to emulatebig endian/little endian, and would have to emulate SSE too... windows on x86, and all the ecosystem of software has evolved and gained much in complexity in the past decade. wabi today would be a dead end even before having started
        • mono for solaris sparc

          it does seem to exist. i don't know if the mono 2.0 is available, but it will be eventually, and open source, so anyone can contribute the port
        • No additional complexity in C# ?

          It seems to me that .NET / C# added all (cool) features coming from other languages so .NET pundits could say: we have partial classes and partial methods, we have runtime generics, we have covariance and contravariance, we have (almost) macros, we have lambda expressions, we have anonymous types, we have LINQ, we have forms, we have WPF, etc... Sorry but it does not seems to me to be particularly simpler than Java, or particularly simple at all ;-)
          • Perhaps

            Perhaps you should quit programming then.

            Most of what you mention is either not so complex (what's complex about splitting a class across two files?), or internally complex, but makes life easier for the programmer - for example generics.

            WPF is probably the most complex thing in your list, but it's not a language feature, its a class library.

            I'll take delegates and lambda expressions any day over the clunky verbose way Java makes you define an entire class just to handle one event.
  • The nature of the failure

    It was a failure because of the commodity pricing of intel servers, and people weren't interested in more expensive hardware (at least in that part of the market).

    Dave Cutler famously hates x86, and NT was designed to be portable. It shouldn't be that hard to get Windows Server running again (I'm not suggesting Window client is made to run there).

    As for Sun, they already sell x86 servers with Windows, and how much do they charge for Solaris anyway?

    A commodity market based on something other than x86/x64 would a very good thing for everyone but Intel, but perhaps that a different topic.