SCO, Linux, and Microsoft in the History of OS: 2000s
August 22nd, 2007
Daniel Eran Dilger
[continued from part four: SCO, Linux, and Microsoft in the History of OS: 1990s]
Apple’s Reversal of Fortune, Outlook: 1990 vs 2000.
Apple started the 1990s with seemingly bright prospects and plaudits from the press, but in reality, its management had aimed the company’s future into the ground. While often described as a “Golden Age” for Apple, the early 90s were actually the beginning of an end.
The Mac hardware architecture and its operating system software were both aging and needed replacements, the company’s marketing had created a branding perception of overpriced equipment that pushed consumers to look at PCs. Since the company had spun off its application software to Claris, it was now more dependent upon third party developers for support, but those developers had increasingly fewer reasons to support Apple.
Ten years later in 2000, Apple was regarded by the press as beleaguered and increasingly irrelevant. However, an infusion of new management and technology from NeXT had actually targeted the company’s future at the stars. The media just hadn’t discovered that yet.
Mac hardware was competitive, and Apple’s new Mac OS X operating system was nearing release. Apple’s brand had improved with the introduction of the popular iMac and iBook, and new Mac-only software applications were being delivered in rapid succession to give potential buyers a reason to chose Macs.
Apple’s NeXT Problem: 1997.
Apple labored for nearly a half decade in its efforts to deliver the technology acquired from NeXT in the final days of 1996. The original plan had been to simply replace the Mac OS with a version of NeXTSTEP ported to the PowerPC Mac hardware and dressed up with a Mac-like appearance.
The hard work had largely already been done, as NeXT had started plans to build a NeXT RISC Workstation using dual PowerPC processors shortly before it dropped out of the hardware market in 1993. Apple finished the port within six months and delivered a developer preview at the 1997 WWDC, under the code name Rhapsody.
In addition to giving Apple a new operating system for the Mac, Rhapsody could also run on standard Intel x86 PC hardware. Additionally, a runtime called the Yellow Box allowed Rhapsody applications to also run on Windows. With a little work, other platforms including Sun Solaris could also run Yellow Box applications.
The Death of Yellow Box: 1998 – 1999.
This strategy came crashing down when major Mac developers–principally Microsoft, Macromedia, and Adobe–refused to reengineer their applications to run on Yellow Box and demanded that Apple figure out a way to transition their apps to continue to run on its new operating system.
That resulted in Apple having to build a transition architecture to get developers on board the new Mac platform. Since that took years to build, Apple also had to retrofit the old Mac System 7 at the same time, so it would have something to sell on its existing Macs. It released an initial update as Mac OS 8, and second as Mac OS 9.
That strategy was laid out to lead into Mac OS X, a reworking of NeXTSTEP that added a cleaned up version of the primary classic Mac OS APIs, which Apple began calling the Blue Box and then Carbon.
Portions of the Yellow Box APIs became Cocoa, but such extensive changes were made that Cocoa and Yellow Box ended up as very different things.
Going Open Source For Fun and Profit: 1999 – 2002.
Apple also had to modernize the core OS of NeXTSTEP, which hadn’t been in active development since 1994. Fortunately, the three major open source groups maintaining forks of BSD–FreeBSD, NetBSD, and OpenBSD–had made significant improvements that Apple could incorporate.
Apple also released its unique implementation of the Mach/BSD hybrid core OS of Mac OS X as the Darwin open source project. Since BSD had been freed after Novell’s acquisition of AT&T’s Unix System Labs, Apple could clean out the dependancies on Unix licensing. In addition, other proprietary licensing was removed, including Adobe’s Display PostScript imaging layer.
These changes not only modernized the code and cleared the way for Apple to embrace more open development, but also allowed it to dramatically lower the price. NeXTSTEP had cost $795 per user, or $2790 with a full set of developer tools.
Mac OS X would end up costing $129 with full developer tools. Apple later threw in a deployment license for WebObjects, which had originally cost $50,000.
Apple Takes on Microsoft: 2002 – 2007.
Mac OS X successfully merged a modernized update of the core Unix OS of NeXTSTEP and its Enterprise worthy objected oriented development tools with Apple’s existing consumer Mac market, hardware manufacturing operations, established sales channels, and unique technologies including QuickTime and ColorSync.
The new infusion of technology enabled Apple to rapidly outpace Microsoft. After delivering a fifth version of NT as Windows 2000, and subsequently rolling in support for consumer hardware and gaming with the 2001 update Windows XP, Microsoft embarked on five years of abortive failure in releasing its next major update of NT 6.0, code named Longhorn and eventually shipping as Windows Vista.
In the meantime, Apple released its first mainstream version of Mac OS X 10.2 Jaguar in 2002, followed by 10.3 Panther in 2003 and 10.4 Tiger in 2005. Being built upon a Unix foundation also allowed Apple to provide X11 compatibility, making Macs attractive to users in higher education and within science and technology markets.
At the same time, Apple began targeting creative businesses with new applications, building upon its established markets in video and broadcasting and moving into music and photography.
It also launched a series of consumer applications to make the Mac platform attractive to power users working with music, photos, and video. These iLife apps distinguished the Mac as a ready to use product.
Caledera’s OpenLinux: The Linux “Mac OS X” That Failed: 1996 – 2002.
Apple wasn’t the only company working to take on Microsoft. After Ray Noorda’s plans to build a rival to Microsoft’s Windows and Office businesses using Unix and WordPerfect were abruptly ended by his ouster from Novell, he began an independent second attempt in his 70s using Caldera Systems.
Noorda partnered with Ransom Love, a Novell marketing executive, to develop a cross platform, Linux-based desktop that could rival Windows. Caldera had acquired assets related to DR-DOS from Novell in 1996, and used these to win a settlement against Microsoft for its illegal actions against Digital Research from 1981 to 1995.
Caldera wasn’t just a litigation company, however. From 1996 to 2002, Caldera had worked to deliver OpenLinux as a common application platform that developers could use to build Linux applications. The problem was that there was a very small installed client base for Linux in the Enterprise, and no significant desktop applications for Linux. Noorda hoped to solve this catch-22 by working to:
- port the Corel WordPerfect and CorelDraw apps to Linux.
- migrate the existing, significant installed base of commercial Unix users to Linux.
Users + Applications = Platform.
This strategy was similar Apple’s plan of the same time period to migrate its existing classic Mac installed base to the new NeXTSTEP-based Mac OS X, while also leveraging Unix and open source software so there would be both enough applications and users to make the new platform viable.
After winning its settlement against Microsoft in 2000, the relatively tiny Caldera had the cash to buy up the much larger Santa Cruz Operation, the company that now held distribution rights to the commercial Unix that Noorda had acquired from AT&T while at Novell, only to subsequently lose after he was dismissed from Novell.
Caldera acquired the Santa Cruz Operation and began efforts to move its commercial Unix user base to Linux. Those users included both SCO Unix, originally derived from Microsoft’s Xenix, and UnixWare, the more modern but still dated AT&T Unix.
While Linux was already far beyond the capabilities of the then stagnant commercial Unix variants, it was still a tough sell to move entrenched Unix users to Linux. Linux was fractured into different distros just as Unix had been in the 90s, complicating maintenance and development. That was particularly an issue for commercial Unix users who expected more conservative stability from their vendor.
Caldera’s Failed United Linux: 2002.
Two years later, in 2002, Caldera formed United Linux in a partnership with SuSE Linux, Turbolinux, and Conectiva. The goal was to develop a Linux platform targeted to Enterprise use, primarily to compete against the established Red Hat Linux distro.
The United Linux partners planned to develop a common platform called the Linux Standard Base, removing some of the uncertainty involved with the differing Linux distros; they essentially decided to standardize on SuSE’s technically superior implementation of its Linux distro and co-market it in their respective locations:
- Caldera in North America
- SuSE in Europe
- Turbolinux in Asia
- Conectiva in South America
Caldera’s commitment to United Linux began to fall apart under the challenge of migrating its Unix customers to Linux while also changing the Linux target from OpenLinux to SuSE. In parallel, Caldera worked to create a Linux/Unix hybrid to sell to its commercial Unix customers acquired from the Santa Cruz Operation.
Linux didn’t offer the scalability and familiarity that UnixWare customers wanted, so Caldera paired the work it had done on OpenLinux with the existing UnixWare kernel to deliver OpenUnix 8. It was essentially standard commercial Unix with a Linux compatibility layer.
[OpenUNIX 8 - Unix Review]
Caldera’s War on Linux: 2002 – 2003.
The company moved even further from Linux when Caldera brought on McBride as CEO in the middle of 2002. McBride rebranded Caldera as The SCO Group, recognizing that the majority of the company’s business was made up of customers still using legacy commercial Unix.
Noorda’s plans to migrate those users to Linux had largely failed, and the United Linux effort intended to move SCO Unix customers to SuSe Linux, which made little business sense for SCO’s investors. Noorda, now approaching 80, retired and Caldera’s corporate personality changed dramatically.
Under McBride, The SCO Group redefined itself as the standard of all commercial Unix. This took a weird turn when McBride announced that Linux improperly violated its copyright by incorporating commercial Unix code. Defenders of Linux maintained that all the code in Linux was independently implemented. Everyone associated with Linux developments at SCO began to jump ship, including Noorda and Love.
Somewhat ironically, United Linux folded in 2003 when Novell bought SuSE to adopt it as its new, ongoing Linux strategy, effectively pitch hitting for Noorda in a fourth swing at creating a new platform to rival Windows:
- Noorda’s 1993 Novell Unix.
- Noorda’s 1996 Caldera OpenLinux.
- Noorda’s 2002 United Linux.
- Novell’s current SuSE Linux.
Caldera vs Microsoft and SCO vs Linux.
SCO’s assault on Linux was characterized by Daniel Lyons and other observers as a simple replay of the litigation Caldera led against Microsoft related to DR-DOS. This wasn’t accurate however.
The difference between SCO’s lawsuit against Linux and Caldera’s earlier lawsuit against Microsoft was that Caldera had a smoking gun, a dead body, a motive, and an opportunity all documented against Microsoft. There was no case to argue; DR-DOS had its hands tied and was publicly shot in the face by Microsoft in broad daylight.
SCO had nothing more than accusations against Linux, and refused to offer any reason to believe any of its claims. Over the next five years, SCO argued that Linux users owed it royalty payments for borrowing lines of code that it refused to present any evidence of, and warned companies that any use of Linux could expose them to expensive legal problems.
DR-DOS Settlement Reverts to Fund Microsoft’s War on Linux: 2003 – 2007.
In 2000, Microsoft had no problem paying out hundreds of millions to Caldera to cover up its ugly history related to DR-DOS. Rather conveniently for Microsoft, just two years later Caldera’s Microsoft settlement had been converted into SCO, an anti-Linux litigation company that now served Microsoft’s own interests.
SCO needed more money to maintain its years of war on Linux, so Microsoft donated millions of dollars in “Unix licensing” to help SCO continue its litigation against Microsoft’s competitors, including IBM and Novell.
SCO initially sued IBM for a billion dollars, alleging that it had copied SCO Unix code into AIX and Linux while working with the old SCO on Project Monterey, a late 90s effort to port Unix to Intel’s Itanium IA-64 fiasco. SCO also claimed IBM’s own enhancements to AIX could not be ported to Linux because they were “derivatives of commercial Unix,” simply because they were originally built for a Unix platform.
Novell was involved because after SCO sued IBM over copyright infringement, Novell announced that it had never actually sold the copyright to Unix to SCO, and that SCO’s case was without merit. Novell insisted that it had only ever sold SCO the rights to distribute UnixWare. Outraged SCO lawyers then sued Novell for slander of title.
That claim was shot down last week when Judge Dale Kimball handed down a ruling that found Novell did indeed own the copyright to Unix and UnixWare. That in turn deflated SCO’s case against Linux, as it has no basis to claim any copyright infringements at all.
The Futile Fight Against the Future.
Rather than trying to adapt and change for the future, SCO hoped to prevent it from happening, just as IBM once tried to hold back personal computing, just as DOS users in the 80s tried to hold back graphical computing, and just as Microsoft is working to hold back free and open source software computing today.
Efforts to fight the future through threats of litigation similarly promise to ensnare Microsoft in a war it can’t win as it seeks to maintain its illicitly acquired and scandalously maintained monopoly position in PC software.
Like reading RoughlyDrafted? Share articles with your friends, link from your blog, and subscribe to my podcast! Submit to Reddit or Slashdot, or consider making a small donation supporting this site. Thanks!