Mon 29 Sep 2008
FREE SOFTWARE upstart and Unix clone Linux is on track to see off the proprietary flavours of Unix, opined the Linux Foundation Executive Director Jim Zemlin in a feature article at Infoworld on Wednesday.
Zemlin leveled his predictions of inexorable decline and doom at Sun Solaris in particular.
“The future is Linux and Microsoft Windows,” said Zemlin. “It is not Unix or Solaris.”
That statement strikes us as odd, because Linux is an architectural descendant of Unix that runs software easily ported between the two environments, and both are far different from Microsoft’s consumer oriented Windows operating systems. But his argument went further.
Zemlin has the attitude that Sun and Solaris should get out of the way of Linux takeup. He derided Solaris as a legacy operating system that’s not gaining in customer deployments.
He also claimed that computer hardware manufacturers don’t see much future for Solaris and that Sun represents an operating systems software vendor facing financial difficulty.
Zemlin pointed out that Linux leads in new deployments on x86-based systems, where the lower cost of commodity x86 hardware gives it an advantage versus lower volume big tin servers running proprietary Unix.
He said Sun’s strength has been in commercial software applications such as Enterprise Resource Planning systems with seven to 20-year life cycles. “What’s starting to happen is those life cycles are starting to be completed,” he said, adding that those applications are being replaced on systems running Linux.
Zemlin also claimed that the industry trend toward web-based applications is accelerating Linux adoption among system developers, saying, “You can’t really talk to any web-based application company these days that’s not using Linux.”
Linux is also less expensive for users to run than proprietary Unix, he noted, suggesting that Sun should simply move to Linux instead of continuing to develop and support Solaris. He also implicated the two other proprietary Unix brands, IBM’s AIX and Hewlett-Packard’s HP-UX. “It’s certainly true that Unix is on the decline,” he said.
He could have added that both IBM and HP have both been strong supporters of Linux for several years now, even though both companies continue to support their slowly-declining base of legacy Unix customers. That Sun hasn’t joined them might be what exercises him.
Zemlin warmed to the topic of user migration from Unix to Linux. “Customers are pretty aware that Unix is a more expensive legacy architecture. They continue to support it because they don’t want to change their legacy apps over to a new platform because of the [transition] costs.”
Well… it’s fairly easy to show quick recovery of transition costs from savings due to lower hardware, software, support and operating costs, though admittedly there are always up-front expenses and risks associated with such migrations. But lower ongoing costs rule.
Zemlin went on, “But they know now they eventually need to do it because Unix just doesn’t have the combined might of all the different organisations and individuals that are developing Linux.”
His implication, we think, is that even if Solaris is still somewhat more sophisticated than Linux in certain areas, Linux is developing so fast that it will surely catch and overtake it.
There’s such weight behind Linux system development now, not just in the kernel itself but also at all levels, that it’s hard to argue with this proposition. Proprietary Unix is doomed.
There’s more discussion of Solaris Dtrace and Zettabyte File System (ZFS) and how those aren’t really big advantages for Solaris, according to Zemlin, plus his wish that Sun would open source those technologies and make them available within Linux.
Naturally, Sun Microsystems doesn’t view Solaris’s future prospects inquite the same way.
The Linux Foundation and Sun will likely continue to disagree about the relative places of both Solaris and Linux, right up until Sun moves further towards open source and joins the Linux Foundation, which we believe is something that is likely to happen eventually. µ