Caldera OpenLinux

From Wikipedia, the free encyclopedia

The neutrality of this article is disputed.
Please see the discussion on the talk page.

Caldera OpenLinux is a defunct Linux distribution that was created by the former Caldera Systems (now SCO Group) corporation. It was the early "business oriented distribution" and foreshadowed the direction of developments that came to most other distributions and the GNU/Linux community generally.

Contents

[edit] Novell and Corsair

Corsair was a project founded inside of the Novell corporation. To quote the press release:

Corsair, as a new user interface for Netware. Corsair is what CEO Robert Frankenberg likes to call a "net top", a user-friendly interface for navigating around computer networks — and the Internet — that is icon-based, uses 3D graphics, and photographic-quality images, a more sophisticated approach than the cartoon figures in Microsoft's Bob interface.

This was being developed by a group inside of Novell called the Advanced Technology Group. They wanted an Internet desktop and were doing research on how to better and more easily integrate and manage network access for users. Windows networking support was terrible[1], the internet was dominated by UNIX based operating systems. Relative to their needs, Novell deemed the Unixes of the day [2]:

  • Too hardware intensive
  • Too large
  • Charged too much in license fees

This group became convinced that Linux offered the best possible answer for the OS component. There were many other components as well, of particular interest were

On April 5th 1994 the Board of Novell brought in Robert Frankenberg, the general manager of Hewlett-Packard Personal Information Products Group to replace Ray Noorda as CEO of Novell. Novell's stock price had performed poorly recently due to flagging growth. There were 3 divisions within Novell:

  • The Network division growing at 19% per year
  • The UNIX business division which was flat
  • The Desktop Applications division which was shrinking at a rate of $400 million per year

Frankenberg's initiative was to refocus the company on networking and networking services. In terms of the Corsair that meant shedding most of the pieces. Negotiations started which would eventually lead to Wordperfect being sold off to Corel. The advanced technology group was disbanded which shut down Willow and the OS project. Ferret was in line with the new direction and this component was kept within Novell. Ray Noorda had founded a venture capital investment group called the Canopy Group two years earlier. He felt there was substantial promise in both the OS project and the Willow project. He created two companies, to continue the work started at Novell. The OS company was called Caldera System, and the API company Willows Software.

Noorda's early vision for Caldera was to create an IPX based version of Linux which would license the key components, resell this technology back to Novell to continue the "Internet Desktop", to quote Noorda, "Caldera Network Desktop for Internet Access, a commercial grade implementation of Linux with built-in IPX, Windows for Workgroup and Internet protocol support and support for DOS/Windows/Unix applications". In effect Caldera started life as an outsourcing project for Novell. Caldera started with ten employees and most were from Novell: Bryan Sparks, founder/president (Novell); Bryce Burns, chief operations officer (Novell); Ransom Love, VP marketing (Novell); Greg Page, VP engineering (Bell Labs, AT&T); Craig Bradley, VP Sales (Lotus, Word Perfect).

[edit] Caldera Network Desktop

Back of Network Desktop 1 box
Back of Network Desktop 1 box

At this point Ransom Love and Noorda took note of the technologies that Caldera put together.

More than just a component for Novell, Caldera has assembled the components needed to create a VAR platform. Caldera faced a chicken and egg problem. OEM VAR applications often depended crucially on other company's commercial applications. Since these other applications hadn't been ported to Linux yet they couldn't meaningfully port their own applications. Caldera responded by creating a binary applications package which allowed Linux to run Unixware and OpenServer applications, the Linux ABI project, and assisting SCO in creating the LKP(Linux Kernel Personalities). To quote Love, "We worked on Linux Kernel Personalities to bring Linux application compatibility to SCO Unix (formerly UnixWare) and OpenServer. The idea was to enable developers to write for both Unix and Linux with a common Application Programming Interface (API) and common Application Binary Interface (ABI). That way developers didn't have to work so hard, and Unix users, the client base we inherited from SCO, could run Linux applications. We were no longer thinking about mixing code; we were trying to create a common development environment. We were trying to keep the Unix and Linux kernels separate, while tying them to common APIs and ABIs."

Caldera also supported Alan Cox in his work on SMP, "Caldera bought the hardware... the dual P90 board + processors was not exactly cheap. The board btw is alive and well and currently owned by Dave Jones"[3] Intel was supportive of Caldera's strategy. If Linux destroyed the Unix base on Intel then Sun wouldn't have a low-end Unix path. This point becomes more interesting in light of SCO's litigation 8 years later.

It is worth commenting however that:

  • IBM was not the company involved
  • The company most directly involved is the company that later became the SCO group

One might conclude SCO is suing IBM for its own acts.

Here is a review from Linux Journal of a late beta of Linux Desktop 1.0, and the original press release with more information.

During 1996 Caldera continued to be a valuable player, for example on May 23, 1996, at the Linux Kongress in Berlin, Germany, Caldera announced its plans to obtain POSIX and FIPS Certifications and the X/Open brand for UNIX 95 and XPG4 BASE 95 for the Linux operating system kernel.

[edit] Caldera Open Linux

Caldera Logo
Caldera Logo

By 1997 Caldera had taken on the form that it would be most remembered for. Caldera had switched over to the high end Linux product,

In 1995 for example when XFree86 was still very hard to configure and unreliable on most chip sets, Caldera had shipped with MetroLink's Motif and XI Graphic's accelerated X.

Listing the software for any particular year for a dead product is pointless but over the next 5 years they offered additional commercial extensions to Linux (including but not exclusively):

  • They licensed Sun's Wabi (Windows Application Binary Interface) to allow people to run Windows applications under Linux.
  • From Novell and later Corel they shipped with Linux versions of Word Perfect and CorelDRAW.
  • They felt an integrated office suite was essential and shipped with the integrated StarOffice (the version acquired by Sun not the version produced by Sun).
  • Since many of their customers used a dual boot setup and FIPS was unreliable they shipped with PartitionMagic to allow their customers to non destructively repartition their disk.
  • They continued to bundle in their Netware Server and Netware clients for Linux.
  • They worked with IBM and produced the first Linux distribution which was DB2 compatible. They also shipped with a demo of DB2.
  • They worked with the Oracle corporation and were the target platform for the Linux port of the Oracle database.

In addition to other people's apps they created many Linux extensions to fill voids where no other commercial company was.

  • They began working on a Linux equivalent of replacing Microsoft Exchange Server and Microsoft Outlook that would eventually become Volution Messaging Server. Volution Messaging Server which was a replacement for exchange server integration with Microsoft's Outlook and offers calendaring/scheduling options with shared busy/free information, SSL support for e-mail and easy configuration, Here is a review of what the product became.
  • They created the first fully graphical installer for Linux. Lizard
  • They invented browser based UNIX system administration and created the webmin project.
  • To help port DOS applications they purchased DR-DOS and used the source code to create OpenDOS.
  • They created a full featured GUI system administration tool COAS. The tool was:
  • They were the first distribution to be pick a specific GUI, and oriented themselves as a KDE distribution long before any other major Linux distribution was doing anything similar. Mandrake followed soon thereafter.
  • They started the Blackdown Java project.
  • The early on saw the role for Linux in the embedded market space and founded the Embedix distribution, which was an early attempt to create a real time Linux. Other Linux vendors generally went for Linux APIs on top of other people's RTOS. For example RedHat was using Linux apis on top of ECOS in its ELIX project.

Politically they also drove Linux in a commercial direction.

By the end Caldera offered 3 versions:

Henry Gleitman had commented that while Freud's theory itself was wrong, "Freud provided a goal for posterity. He showed us the kinds of questions that we have to answer before we can claim to have a full theory of the human personality." The analogous point can be made of Caldera OpenLinux, OpenLinux was not a Microsoft killer, but it showed the Linux community what would be required to create a mainstream desktop OS out of the Linux kernel. In many ways the last 10 years of desktop progress has been to successfully implement what Caldera was attempting to do with the tools they had available. Their technique for this was to utilize commercial software to fill in the largest gaps. This made their product a "value add" and thus they could charge for it, and at the same time it made them the most advanced distribution available.

[edit] Lineo and Marketing

Lineo Logo
Lineo Logo

In July 1999 Caldera decides on a major refocus, on the embedded side where Caldera's technologies that were way ahead, were owned by them. These were:

  • Rt-Control provided UCLinux, a version of Linux for microcontrollers such as the Motorola 68k/ColdFire line, i960, ARM7, and ETRAX CRIS chips. Because these chips have no MMU (thus cannot provide multi-tasking capabilities), uClinux is able to run full-featured in as little as 150 KB of RAM with a 1 MB ROM chip.
  • FirePlug focuses on specialized Linux-based project such as their Linux firewall. They build these on the ThinLinux product, which runs in as little as 2 MB of disk/flash storage and 8 MB RAM.
  • Embedix, Lineo's flagship product, runs a complete multitasking, networked Linux operating system in 2 MB of ROM/flash and 4 MB of RAM.
  • Embedix Browser was a fully graphical internet browser for embedded systems, this was in essence the Zenotropix browser.
  • Embedix SDK which was the Zentropix SDK.

This combination of technologies allowed Caldera to offer a full Linux operating system with a graphical browser that could run off a floppy disk. More importantly the product was unique, and this came from the fact that Caldera's view on the Linux embedded market was differed from other vendors. All the other vendors believed that Linux was heavily fragmented and that the solution was to offer Linux features for real time OSes, that is a Linux API for some other OSes. Red Hat with its EL/IX created a kernel independent framework (API) which allowed some Linux software to run on the ECOS kernel. Caldera did not agree with this assessment and believed the API was offered far more advantages and allowed for a fully hardened system, that is Caldera utilized a custom Linux kernel. As Bryan Sparks said, "Through the six companies we've acquired, Lineo has been able to extend the same (Linux) technology across multiple chip architectures, and we've also added real-time capabilities. Our acquisitions have given us a broader breadth of Linux support, from very small microcontrollers, through traditional platforms like x86, and up to high end, high availability systems. We see our strategy as a 'breadth strategy': moving a common api (i.e. the Linux api) in multiple directions, to gain broader market share."

And thus Bryan Sparks CEO or Caldera Systems, split off the embedded Linux company Lineo as a technology company under his own direction. The desktop company became Caldera International under the direction of Ransom Love. The focus for the desktop company became mainly marketing (Ransom Love was promoted to CEO from VP of Marketing) and business relationships. There were several reasons for this. The first was that Caldera had won a two hundred and fifty million dollar lawsuit against Microsoft for DR-DOS and was flush with cash. Secondly, while the Caldera distribution was good, its primary advantages were the use of technologies not owned by Caldera and thus if Caldera were successful its success could (from a technical standpoint) be imitated, by Red Hat, SUSE, TurboLinux, etc... For years Caldera had been competing directly with SCO Unix, but by 1997 Linux outperformed SCO in almost every respect. But making the choice to switch from SCO to Caldera was not a "no-brainer" for companies because that also meant a switch of vendors and support organizations. Caldera's SCO acquisition was aimed at eliminating this problem. That is Caldera International's corporate direction became: combine SCO's distribution, marketing and VAR arm with LAMP, and use Project Monterey to develop a 64-bit strategy. In short what SCO offered was:

  • A strong list of business clients.
  • Higher compatibility between SCO and Linux than any other Unix/Linux combination, mainly as a result of Caldera's long standing SCO focus that created products like ABI and thus resulted in ports of SCO code to Linux
  • A good back-office and database solution while Linux specialized in networking (LAMP) and client desktop, a very appealing combination in challenging Sun and Microsoft
  • A global infrastructure (presence in about 80 countries), Caldera was domestic
  • Thousands of business applications targeted to vertical markets
  • Some of the 3rd party components needed to get HPUX, AIX, Solaris 3rd party Java applications ported to Linux

From a technical standpoint however Caldera Open Linux really shined during the Ransom Love years. Their commercial bundling solution continued to work. They had a powerful low bug (by Linux standards) distribution that worked well on a wide range of hardware. They charged a great deal relative to other distribution and were able to generate a very strong profit. Red Hat pulled way ahead of them in terms of US sales and on the global sales front they trailed SUSE and TurboLinux as well, but financially due to the DR-DOS settlement they were the strongest of all the Linux distributions.

[edit] United Linux

United Linux mascot
United Linux mascot

Caldera quickly found itself in a classic business problem where the interests of the existing business conflicted with their growth model. SCO was a much larger company than Caldera International had been (the DR-DOS settlement had been what made the buyout possible), and in fact of the $71m of revenue 90% was from the SCO side of the business. Moreover, Caldera costs $2 in marketing to generate a $1 in sales, SCO was mature and sold itself (mainly to repeat customer). The VAR relationship was even more problematic. Caldera had always sold the "Linux is SCO but better" model and had done everything possible to make the transition from SCO to Caldera relatively seamless. Each of the 14,000 SCO resellers made much more from each SCO sale than from sales of Caldera so they were not anxious to move existing customers from SCO to Linux; and even those that were supportive of Linux saw no strong value add for Caldera and often sold Red Hat Enterprise Linux. Caldera had two businesses in direct competition one which was a shrinking but still profitable UNIX business, the other a rapidly growing business that was still hemorrhaging money.

The most logical solution was to establish Caldera as the premier Linux brand. Without the threat from Red Hat, transitioning resellers from SCO to Caldera would be much easier. With this in mind Ransom Love formed an alliance of large business oriented Linux distributions which utilized the KDE desktop, called United Linux:

  • Caldera
  • SUSE Linux, one of the largest distributions in Europe, and the most business oriented
  • Turbolinux, the largest distribution in Asia
  • Conectiva, the premier business oriented Linux distribution in South America

As filings from Novell in the SCO Group SCO v. Novell lawsuit show us this was more than simply a marketing gimmick:

In particular the United Linux members agreed that each member would have an irrevocable, perpetual, and worldwide license to use and unlimitedly exploit any intellectual property rights of the other members in the UnitedLinux Software, which would be transferred to the LLC for this very purpose...

Business responded favorable to the movement: IBM and AMD quickly formed partnerships, the Linux Professional Institute adopted United Linux as their standard distribution for training. In many ways it could be argued (and United Linux itself argues this) that with United Linux November 2002 release Ransom Love accomplished the goal that Caldera had been working for during the last 10 years. For the first time there was a Linux distribution with:

  • Global scope
  • Global support at the VAR, OEM and distribution level
  • A full training organization
  • Some governmental buy-in
  • Support from major corporations
  • Enterprise applications like Oracle supported out of the box
  • An actual production GUI that ran well on a variety of hardware

SUSE had the engineering as it had continued to maintain a large technical staff, Caldera had the global support organization and Turbo Linux and Conectiva brought with growth potential into less flooded markets. This merger was so successful that Love and Sparks could claim vindication that year when Novell reversed the Frankenberg decision and brought United Linux engineering talent back into the fold with the acquisition of SUSE.

United Linux was rejected by the broader Linux community however. The use of per seat licensing was the most highly controversial move by United Linux for example. More importantly by the time United Linux was released Caldera was already dead. Darl McBride had become CEO of Caldera International and the focus had shifted away from Linux.

As an incidental Caldera at this point released a Caldera "Linux distribution" with the OpenUNIX 8 kernel instead of the Linux kernel. Unix has TLI and STREAMS support which makes writing drivers easier. This product supports Richard Stallman's point that Linux is merely a kernel for the GNU system. Caldera proved this by replacing the kernel and yet not having to change much else on a full featured desktop and server "Linux".

[edit] Copyright infringement allegations

Further information: SCO-Linux_controversies

In 2002, the Caldera board of directors, including Ralph Yarro, brought in the CEO of Franklin-Covey Darl McBride. Almost immediately he saw the value of Caldera as being primarily the value of SCO[citation needed]. The company was renamed The SCO Group. Ransom Love was reassigned to work exclusively on United Linux. After he completed this, he left the company to join Progeny Linux Systems which was aiming to create a professional Debian. He remains there in the capacity of a board member and advisor[2].

McBride began to focus on SCO's copyrights. One of McBride's first acts as CEO was to collect $600,000 in back licensing fees that were owed due to Caldera. He cleaned up various Linux related licensing issues allowing for a new round of financing [3]. Soon thereafter he made strong accusations that Linux had infringed copyrights SCO held on UNIX; they claimed to have purchased these copyrights from Novell. Novell denied selling them UNIX, prompting them to sue for slander of title. SCO also initiated lawsuits against IBM and AutoZone, alleging copyright infringements through the use or distribution of Linux; none of these lawsuits have been resolved. SCO has created a division, SCOSource, that owns and licenses their intellectual property; a desktop license is $199.

[edit] References

  1. ^ Microsoft itself has indicated this. In their history of Windows [1] they specifically indicate that prior to Windows for Workgroups Windows–based PCs were not network-aware and were not a meaningful part of the emerging client/server computing evolution.
  2. ^ http://www.progeny.com/about/board.htm#love
  3. ^ http://news.com.com/2100-1001-939881.html

[edit] External links

  • Groklaw has an extended discussion of the Linux ABI
  • The Japanese subsidiary of Lineo (United System Engineers) is still alive.
  • There is a focus in this article on Caldera vs. other distributions with a particular focus on RedHat. The reader may like to view a timeline for Debian to enhance understanding of what was going on in the broader GNU/Linux community at the times discussed in this article.
  • An advertisement / white-paper from March 2001 comparing RedHat to Caldera. This pushes the "product not a distribution" philosophy [5]
  • Ralf Flaxa's LST [6] was a major contributor for the original Caldera Desktop [7]