I wrote this paper as an assigned topic for Ken Rufo’s “Evolution and Trends in Digital Media” class for the Master of Communication, Digital Media program. I chose this one because I had purchased a PowerComputing Mac clone during the period when they shipped with BeOS as well as MacOS, and I had followed the long saga leading to the eventual purchase of NeXT by Apple with considerable interest. The paper could use another editorial pass, as my last edit for length introduced some choppiness, but I think it’s a good read and a story that you don’t have to be too much of a geek to find interesting.
“Hammerheads”
Computer Operating Systems provoke emotional reactions in people, and sometimes these reactions turn into heated disagreements. While I’ve never seen an OS argument come to blows, I have no doubt that it’s happened.
Since Operating Systems are just tools for accessing the instructions contained in computer hardware, this can seem irrational. It’s like getting emotional over the difference between a carpenter’s claw hammer and a metalworker’s ball peen hammer. They both are used for hitting things, but they hit things in different ways and are used for different jobs. I have no doubt that hammer arguments also have occurred, and come to very painful blows as well.
This passion makes some sense if we think of tools as media. If media are extensions of ourselves, then the Operating System interfaces and hammers we choose reflect how we present ourselves to the world. (McLuhan, 1964)
This idea is contained in the common expression “to a man with a hammer, everything looks like a nail.” What differences are there in the way the world looks to users of Windows, Macintosh OS X, Ubuntu, or any other Operating System?
“Nemesis”
In 1991, the computer business landscape was in the midst of a shift.
In the 1970′s and early 1980′s saw the emergence of successful computer companies selling proprietary hardware, requiring Operating Systems written expressly for them. Significant players included Commodore, Tandy/Radio Shack, Atari, Texas Instruments, Apple, and NeXT Computer.
By 1991, Tandy/Radio Shack and Texas Instruments we no longer selling computers. Commodore and Atari were failing and Apple was headed for a stock price crash, and NeXT Computer was seen as innovative and influential, but not viable.
In contrast, Microsoft was enjoying a remarkable ascendence. By staying out of the hardware business, focusing instead on software, they had been perfectly positioned to benefit as the IBM PC design was copied by numerous companies. This meant their MS-DOS Operating System licensed by IBM would run these clone computers as well.
Among other competitive advantages, avoiding platform integration allowed Microsoft to benefit from price competition between hardware companies, without actually facing significant competition for its own products. Whether it emerged by design or by luck, this proved to be an enormously successful business model for dominating the business user market.
Then, as more people learned to use MS-DOS and early versions of Windows at work, they began buying home computers running these familiar Operating Systems at home as well.
“Health and Knowledge and Wealth and Power”
In 1991 two notable new Operating Systems began development.
The first Linux kernel ran on IBM PC clones. Unlike Windows (or UNIX) it was free and within a year would be distributed with an open-source license. The first public release of Linux was version “0.01″ denoting how incomplete it was after a few short months of part time development by Linus Torvalds in his spare time. (Hasan)
In contrast, BeOS was an Operating System intended to follow the tightly-coupled hardware/software model. Like Linux it was written from the ground up, but unlike Linux it was never released as a work in progress. The first release of BeOS would not occur until 1995, when it would prove to be stable, powerful, and very well-suited for the multimedia computer experience that has become common in the last few years.
Be Inc. founder Jean-Louis Gassée had overseen advanced product development and worldwide marketing at Apple. Previously, as the head of Macintosh development, Gassée had resisted the idea of licensing the Operating System to other hardware manufacturers, believing that hardware control was a key component of Apple’s strength. (Hormby, 2005)
Perhaps because of this belief, the first prototype BeBox running the BeOS used AT&T’s Hobbit chip, rather than the chips from Intel or Motorola used by other computer companies. (Lampert)
When AT&T stopped Hobbit production, Be Inc. needed a replacement platform immediately and found it in the PowerPC chips that Apple had jointly developed with IBM and Motorola to replace the 6800x processor in Macintoshes. Thus, when the first BeBox shipped in 1995, it was very similar to the hardware used in the first Macintosh PowerPC’s released the previous year. All told, around 2000 BeBoxes were sold.
“Everything That Rises Must Converge”
Back at Apple, the idea of increasing market share by licensing the MacOS was regaining traction. However, unlike Microsoft, which had a minimal interest in the hardware Windows ran on, Apple had until recently been the world’s single largest manufacturer of desktop computers. Not wanting to open their lucrative hardware line of business to the cutthroat competition that made IBM PC clone hardware so cheap, they chose a middle path and licensed their proprietary hardware designs and OS to other systems manufacturers.
Be Inc. knew that hardware was only part of Apple’s problem. The Macintosh Operating System, which had had new features haphazardly layered on top of it since its initial release in 1984, had become bloated, unstable, and a major point of user frustration.
A new MacOS project launched in 1989 had struggled from lack of management attention, until by 1994 the situation had become dire enough that it was seen as a crisis. Running MacOS on faster hardware wasn’t going to be enough to put the company back in serious competition.
The new OS project was relaunched and codenamed “Copland” after composer Aaron Copland. Every month, magazines such as Macworld and MacUser provided updates on Copland’s progress. It clearly wasn’t going well.
An initial developer release of Copland was sent to select partners in 1996, and proved to be so unstable as to be unusable. The project was cancelled, and in a surprising break with its history and corporate culture, Apple began looking to outside companies for ideas. (Linzmayer)
BeOS had everything the new MacOS needed to be a modern Operating System: protected memory, multithreading, multitasking, all the geeky details that a user experienced as speed and stability. Negotiations began with Apple to buy Be Inc. but failed, when Apple refused to pay the price Gassée was asking. (Linzmayer)
In December 1996 Apple bought Jobs’ NeXT instead, paying more than twice what Gassée had wanted. Work immediately began on turning the NeXTSTEP OS into what later became released as Macintosh OS X. (Linzmayer)
Looking to get BeOS into people’s hands, in 1997, Be Inc. licensed the BeOS to Macintosh clone maker PowerComputing. BeOS could be installed from a CD that shipped with the clone, allowing the computer to boot into either BeOS or MacOS at startup. User reaction was positive: BeOS was fast and stable, but more than that, the multimedia experience was clearly years ahead of any other OS most people had seen. (Smith, 2007) While the OS X interface owes much to NeXTSTEP, it’s possible to see some influence from BeOS in it as well.
In July 1997 Jobs became CEO of Apple, and, in what might have been a shot at Gassée, declared that the decision to license the MacOS had been made too late to repeat Microsoft’s success. Apple bought out PowerComputing’s license to make Mac clones, which also took the only platform running BeOS off the market.
Seven years after beginning development, and without having yet found a substantial market, Be Inc. ported BeOS to IBM PC clones while continuing support for the PowerPC platform. An embedded version was used in a few dedicated media production systems. Be Inc. sued Microsoft, alleging they had improperly interfered in negotiations with Hitachi to include BeOS with its computers, popular in Europe at the time. The case was settled out of court, years after Be Inc. had suspended operations. (Berniker, 2003)
A version on BeOS intended to run on Internet Appliances — essentially computers capable only of email and Web surfing — was released as BeIA, but the company was chasing a hardware model that would never arrive. In 2000, in an effort to increase interest, BeOS Personal Edition was given away as a download that could be launched from within Windows or Linux.
After 10 years, Be Inc. threw in the towel in 2001. The company was purchased by Palm, and BeOS technology was incorporated into Palm’s Cobalt OS. However, when no licensees adopted Cobalt, it was dropped in favor of Palm’s work on a new Linux-based OS. A few media appliances by companies such as Roland and Tascam still use modified versions of the BeOS or ZETA, an apparently unlicensed update of the original BeOS code.
“Fish Below the Ice”
While undoubtedly very talented, it appears that Gassée was slow to let go of the obsolete tightly-coupled hardware/software model and shaped Be Inc.’s first product offering in a way that made it out of sync with the market he was entering. It’s even conceivable that while at Apple his mindset may have prevented the company from decoupling its hardware and software, a decision that could have cost the company dearly.
In the context of the 80′s, Microsoft’s business model appears as a relatively open one, and that openness is what led to the company’s success. In the context of 1991, Linux came to be the very definition of openness, while BeOS opted for tight control. That control may have produced a great, fully-developed, Operating System while Linux was still a kernel in search of an interface, but over time, the openness of Linux has produced releases like Ubuntu, a free Operating System that comes close to challenging products from Windows and Apple for utility and ease of use.
But in the Internet age, it seems no computer code ever really dies. A group of BeOS fans has been working on “Haiku,” a project to recreate the underlying code, so as to allow the OS legally to be released under an Open Source license, allowing future community development.
While Jean-Louis Gassée might have been looking for BeOS to hit a critical mass of users and explode, the better atomic metaphor here may be that of the half-life. The number of people interested in maintaining an abandoned Operating System decreases fractionally over time. Because the process is one of division instead of subtraction, the number seemingly never quite reaches zero.
“Only Thing That Shines”
When users choose an Operating System, they are choosing a tool to extend themselves. But the OS isn’t just a medium, it’s a tool for making tools — a meta-medium, containing and shaping the experience of all the other media that pass through it. Apple’s marketing department understands this, and I sometimes hate them for it.
I am a cross-platform person, proficient on Mac and Windows, competent on Ubuntu, not unfamiliar with the *nix command line. When I bought a PowerComputing Mac clone dirt cheap after Apple cancelled their license, I even used BeOS to the extent that its lack of practical applications allowed. I can come up with a list of logical, rational reasons for using all these operating systems. Yet when given the option of picking one for general use, I guess the emotional appeal is that I just “think different.”
When users choose a Operating System with marginal software support and a tiny user base, like BeOS in its heyday, is it because they “think differenter”? When BeOS fans dedicate their time to keeping it alive years after it’s become abandonware, what compels them to do so? Is the experience that good, even all these years later when OS X and Windows have UI experiences that look and feel more like BeOS than they do MacOS 7.5 and Windows 95?
Perhaps BeOS might have succeeded if Be Inc. had transmitted this emotional appeal to more people, and all that held them back was the poor timing of their business model, or nefarious maneuvers by other companies. Perhaps not. It’s hard to call BeOS a failure though. They produced an experience that was ahead of its time, coupled to an outdated business model. And in the end, they made a difference.
References
In addition to the sources cited below, some information in this post reflect my own recollection and experiences.
Linzmayer, Owen W. Apple Confidential: The Day They Almost Decided To Put Windows NT On The Mac Instead Of OS X!. Retrieved January 11, 2009, from http://macspeedzone.com/archive/art/con/be.shtml
Berniker, Mark (2003 , September 08). “Microsoft Settles Anti-Trust Charges with Be”. Retrieved January 11, 2009, from http://www.internetnews.com/ent-news/print.php/3073811/
Hasan, Ragib History of Linux. Retrieved January 11, 2009, Web site: https://netfiles.uiuc.edu/rhasan/linux/
Hormby, Tom (2005, June 20). How Jean Louis Gassée Changed the Mac’s Direction. Retrieved January 11, 2009, from Tom Hormby’s Orchard Web site: http://lowendmac.com/orchard/05/jean-louis-gassee-apple.html
Lampert, Andrew Prototype Hobbit BeBox. Retrieved January 11, 2009, from The BeBox Zone Web site: http://www.bebox.nu/images.php?s=images/hobbit
McLuhan, Marshall (1964). Understanding Media: The Extensions of Man.
PalmSource Introduces Palm OS Cobalt, PalmSource press release (Feb 10, 2004). Retrieved January 11, 2009, from http://www.access-company.com/news/press/PalmSource/2004/021004_cobalt.html
Smith, Tony (2007, January 30). BeOS: the Mac OS X might-have-been. Retrieved January 11, 2009, from Register Hardware Web site: http://www.reghardware.co.uk/2007/01/30/forgotten_tech_beos/
Post a Comment