Why Sun won't support Solaris 9 for x86
Summary
Topics
If you go looking for controversy in the computer business, there's a strong possibility that Sun CEO Scott McNealy will turn up in the middle of it. Last week, the Save Solaris Organization published a letter to McNealy as a half page advertisement in the San Jose Mercury News challenging him to a face-to-face meeting over the fate of the Intel x86-based version of Sun's Solaris 9.
Pointing to the fact that the x86 version has traditionally lagged behind the version Sun develops its Sparc-based systems, critics (including Intel CEO Craig Barrett) have argued that Sun has never been serious about x86 support. Case in point: Solaris 9 is already out for Sparc while the x86-version is still stuck at Solaris 8.
But McNealy was right to decline this invitation. As noble as the SSO's goals are, there is little McNealy and Company can do at this point to put Solaris for x86 back on track. The reason Sun has to keep Solaris 9 for x86 on hold is because developing, marketing, launching, and supporting it would be a largely philanthropic effort. There aren't many companies left standing that could afford such philanthropy. Especially Sun.
On the day I started to write this story, Sun's stock missed its 52-week low ($3.43) by two cents. Given the current circumstances, Sun has a fiduciary responsibility to its stockholders not to undertake a project without a clear sense of that project's ROI. Solaris for x86 is just such a project because Sun, for the most part, already gives the Sparc version of Solaris away to its customers.
But, whereas the cost of that philanthropy is subsidized by the sales and service of Sparc-based systems, no such subsidy exists for the x86 version. Because the company ships virtually no x86-based systems, Sun would need to sell the operating system as shrink-wrapped software. Given the going rate for decent graphical multitasking operating systems these days (anywhere from free to $100), any subsequent support contracts, and the relatively few people that would be interested in either, we're not talking about a cash cow here. We're talking about a financial dud.
But all hope is not lost. Sun has a plan, which Sun chief engineer Rob Gingell hinted at during my interview with him. It involves moving Solaris into the open source community where it would be fused with Linux in a way that produces an operating system offering the best of both worlds. Provided the plan can succeed within a reasonable timeframe, Sun could give advocates of an x86-based version what they want without worsening its already resource-constrained situation. Who would have thought that open source might one day get Sun out of a jam?
The plan is not free of obstacles. In addition to the question of whether the Linux community will actually embrace the idea (big if), there's an intellectual property hurdle that has to do with some of Solaris' components. According to Gingell, some parts of Solaris were the product of the Unix community. While Sun is free to include those components in Solaris, it has no legal right to release that intellectual property to the open source community. Gingell seems confident that most of the hurdles can be cleared. But based on the how long intellectual property lawyers take to figure out messes like this, I think six to twelve months is a fair estimate before we see something tangible.
So, picking a fight with McNealy over this doesn't make a whole lot of sense. If you want to bait him, all you really need to do is mention Microsoft.
What are Sun's chances of successfully fusing Solaris to Linux under the guise of open-source? Will the Linux community open its arms or check its pockets? TalkBack below or write to me at david.berlind@cnet.com.
The discussion hasn’t started yet. Why don’t you begin it?
The best of ZDNet, delivered
ZDNet Newsletters
Get the best of ZDNet delivered straight to your inbox
Vendor Showcase
