Should Sun buy Novell?

Sun Microsystems has adopted an ambitious business model that depends upon commodity open-source downloads serving as loss leaders and gateways for hardware and services revenue. According to a report in The Register, however, profits have been hard to come by for Sun, which may have been what scuttled its merger with IBM.

Using Red Hat as a foil, The Register suggests that the way forward for Sun, which has seen its proprietary businesses commoditized, may be to commoditize itself further:

The open source distribution model cannot generate the kind of profits that Sun's shareholders became accustomed to in the dot-com boom, where every deal started out with a Sparc/Solaris server and moved on to Oracle databases....

I can't imagine how Sun's software business--particularly if customers abandon Sparc platforms or Sun has to basically give Solaris support away for free to cover the costs of Sparc chip and server development--can do any better than Red Hat has done on commodity x64 iron. And in the end, the decline in Sparc prices cuts Sun's profits, no matter how it dices and slices the categories and numbers in its presentations, just as the same economic pressures from x64 iron on the one hand and Linux and Windows on the other have done for all proprietary and RISC/Unix vendors.

There is no escaping the pinchers, other that to use the tool yourself. And that means Solaris and x64 are likely Sun's future--and Sparc, for all its great engineering, is probably not.

Let's take this one step further. Maybe it's time to move past Solaris entirely, as the Linux Foundation's Jim Zemlin has argued, and focus Sun's impressive engineering abilities on Linux. Sun could do this by creating its own Linux distribution.

Or it could buy Novell.

Novell has recently seen its Linux business slide, but on balance Suse Linux has been a buoyant force for Novell over the past few years. While I'm not sure Sun can afford to spend much more money to give away free software, I believe the rest of Sun's offerings (software, hardware, and services) will look much more palatable to IT buyers if the conversation starts with Linux, rather than Solaris.

IDC predicts that Linux will grow 21 percent year over year in 2009. I'm guessing Solaris isn't seeing that kind of growth this year...or any time in the future.

Buying Novell would give Sun immediate access to a vibrant partner ecosystem, which is critical: ISVs and IHVs don't want to have to certify for a new Linux distribution.

Again, I know there are plenty of reasons for Sun to not buy Novell, but Suse Linux is an excellent reason why it should double-down on its open-source strategy and fully embrace the operating system to beat in the 21st century:

Linux.

Follow me on Twitter @mjasay.

[lleather]

Novell typically keeps lots of cash around to keep this sort of thing from happening. These rumors have been going on forever regarding Novell.

Seems to me that Novell has the upper hand here due to their obvious foresight regarding the Linux/open source movement.

[gccornell]

microsoft owns part of novell so its not realy opensource any more

[odubtaig]

Correction. MS made a deal with Novell which got Novell a cash injection and a sales partner. MS do not own anything.

[Hep Cat]

Sure, that'll help. After all, when you are drowning in a sea of bad "my little pony" decisions, it's always best to grab an anchor.

[aztracker1]

I'm not sure how that paradigm would line up. Considering Sun is the force behind Java, and Novell is the current owner of Mono. It would be interesting to see the two team up on portability, IKVM, and other methods of getting applications ported from windows, to linux in a cross platform way. Perhaps even seeing the JVM and the Mono CLR see a future point of merging into one Uber-Runtime on the Linux platform.

[dragonbite]

It would be an interesting mix, whether Sun buys Novell, or Novell buys Sun.

Sun could provide the Sparc chip servers that SLES and SLED can be optimized to run on.

The curious contention would be Mono vs Java.

[fennellw]

Sun needs a linux OS that can run all apps that run on Solaris.

[w0rdwarri0r]

The author is totally wrong. Solaris (OpenSolaris) is Sun's differentiator. It has technologies like ZFS and DTrace (and its own kernel, which is many ways is superior to Linux's) that Linux can't use to its GPL license being incompatible with the OpenSolaris license. If Sun becomes a Linux reseller, then they sacrifice a differentiator and a source of profitability. There are tons of companies that can offer support for Linux. Sun (or anyone who it licenses) is the only company that can offer support for Solaris and OpenSolaris.

What Sun should do is implement a better emulation layer for Linux software on OpenSolaris. Brand Z is apparently a little cumbersome to set up.

Besides, Sun has the opportunity to provide the perfect development OS for Java developers. No Linux distro can provide easy updates via its package managers for the latest versions of Java and associated tools. Sun and OpenSolaris can.

[gggg sssss]

close the xompany give any money left to the shareholders. The 1990s are over.

[odubtaig]

Apart from OpenSUSE which include JSE, JEE, the SDKs, etc. in it's repositories.

You don't know much about Linux, do you? Being the only company to support Solaris is ending up like being the only company to support O/S2; no-one cares because no-one's buying.

[ryandsmith]

Running out of things to talk about, Matt? What Sun needs to do is innovate on top of Linux, not let the innovation of Linux run on top of Solaris. Why should Sun buy Linux when the only money to be made on it is in services?

[MikefromDenver]

I think that the "Sun-setting" headlines are overstated. Look at Sun's recent quarters; 4 profitable quarters in a row prior to the start of the recession (the recession hit Sun in their qtr ending March 2008.) The only accurate criticism of Sun has been and continues to be that they haven't balanced their cost structure with their revenues in the post-dotcom world. Investors have already punished Sun for this - look at JAVA's performance over the last 8+ years; $250/share is a distant memory. Assuming the deal with IBM doesn't go through, we'll need to wait until at least end of December 2009 to see if Sun has found the right balance (because revenue during the recession will certainly be down and restructuring talks time.) If they have, then we should expect profits. There are still a lot of customers around the world that want to buy what Sun has - and will continue to pay for innovation. So does that make a $12 billion/year company or a $14 billion/year or who knows? Whatever it is - the costs have to be in line with the revenues. If you bought JAVA this year (in the high $2 - $4 range), you're going to be a happy shareholder. And no, I don't think that Sun should buy Novell - they already have enough to do. :)

[gggg sssss]

If we are at all lucky, JAVA will simply disappear. RIP

[JimCallahan]

Sun plus Novell
Op Sys: Solaris (Unix) plus SUSE
Lang: Java plus Mono (cross platform dot net compatible)
Databases: MySQL
Interface: GNOME (SUSE and Solaris both use Gnome)
VM: Virtual Box and xVM
Both have deals with Microsoft.

Interesting, would only work if market automatically gave larger multiple for Linux.
Jim Callahan
Orlando, FL

[pbouzide]

I like the idea of opening and merging the strengths of Solaris (ZFS notably, but also DTrace and other kernel-level robustness and performance benefits that come with Solaris) with the Linux userland and calling the result (SUSE?) Linux. I could imagine this being successful in the enterprise. I think the comments above do question how realistic this could be though.

[pechter]

I see a different possibility after 20 years in the Unix/Linux Sysadmin business and it's
not just a Linux play.

The obvious reason to do this is not to just do Linux -- but to own the UNIX copyrights and licenses.
After that they can get their SCO SysV buyout blackmail -- er money back which will
then make them the lead in Unix and a major player in Linux.

They can then fully OpenSource SysV... and all of Solaris again... good for PR and it will help reinvigorate that environment.

Could be quite interesting if played right... Gives 'em the leg up they almost had back in the day when they partnered up with AT&T/USL to do SysVRel4.

They'd also get the settlement money from the court when the SCO suit settles down.
Can you say damages anyone?

[dargon19888]

Stallman was wrong. Sun's current situation is proof of it.

There are flaws in the Open Source model and while certain things can be open sourced, other things can not. Sorry for the reality check.

With respect to the concept of Sun purchasing Novell, not an idea grounded in reality. Sun had been shopping itself out to see if it could be acquired. IBM may be taking a breather, but there's always Apple or Cisco that could be a purchaser, in theory ...

[odubtaig]

What reality check? Anyone who's been paying any attention at all knows it's Sun's outdated hardware sales model that's dragging them down while MySQL just keeps brininging new six and seven figure contracts.

You either have to be not paying attention, completely clueless or have an ideological agenda to think that Sun's current situation is down to open sourcing anything,

Just carry on ignoring the success of MySQL and the failure of the Sparc, ignore the massive success of substantially more open companies like Red Hat, forget all the successful businesses centred around F/OSS. One company with a failing hardware division and legacy products it should have dropped years ago is failing despite the success of a newly acquired company with a pseudo-OSS product and Stallman was wrong?

What have you been smoking, son?