The Strange Case of Dr. Open
and Mr. Proprietary
As noted in Chapter 2 of this book, the release of the Altair microcomputer in 1975 heralded the beginning of the modern high-tech industry. But observers of the period also believe there was more to the Altair than just chips; the unit seemed to emit a mysterious elixir that entered the body of computer aficionados worldwide and sparked a strange war of the soul that has raged in the body of the computer geekdom for more than three decades. The war is between those who advocate for free software and open, patentless technology available to all and those who believe in making substantial sums of money from selling proprietary software and the vigorous protection of intellectual property. It’s the Kumbayahs vs. the Capitalists.
Other influences may be responsible for the ongoing struggle. Perhaps Star Trek bears some of the blame. Few in microcomputing hadn’t watched the series, and as Captain Kirk, Mr. Spock, Bones, Scottie, and their innumerable successors went gallivanting through the galaxy, they seemed to have no visible means of financial support. No one in the Star Trek universe wearing green eye shades ever appeared to worry about the propensity of the various casts to blow up what you’d think were undoubtedly very expensive spaceships, given their capabilities of violating the laws of physics, transporting the crew to numerous planets inhabited by women who spent most of their time wearing lingerie, and dodging ray-gun fire from angry races of aliens who kept screaming “kaplok!” (and who also seemed to have no monetary worries). Perhaps the reason for Captain Kirk’s insouciance lay in the fact that everyone in Star Trek had access to what were called “transporters,” magical devices that could be used to whisk you from the spaceship Enterprise to a planet without having to pay a toll. Later in the series’ development, transporters could be used to create chocolate milk shakes, drinks, and even the occasional boyfriend or girlfriend via simple voice commands. And all for free!
Of course, no computer has a Star Trek–like transporter system built into it, but from the standpoint of people interested in obtaining software without forking over monetary compensation, software has something almost as good. That good thing is the “copy” command. And since software, unlike milk shakes, drinks, and boyfriends, is already digitized, just about anyone can execute this wondrous command and enjoy a cornucopia of software in an environment free of the distasteful economic friction of “paying.”
Technology’s interest in the concept of free software was demonstrated almost conterminously with the release of the Altair in the events surrounding the “liberation” of the first BASIC for this pioneering machine. When first available, the Altair had no useful software, and the market was eagerly awaiting the release of Altair BASIC (waiting was something Altairians were very good at doing because Altair maker MITS was legendary for announcing new products it couldn’t deliver, a habit the rest of the industry soon learned to emulate). The product had been developed by a small software firm, Micro-Soft, run by two people no one had ever heard of, Paul Allen and Bill Gates. Micro-Soft had cut a deal with MITS to receive a royalty on every sale of Altair BASIC and was eagerly waiting for a stream of revenue to flow into the tiny firm’s coffers upon the official release of the new product to a marketer eager to buy it.
Unfortunately for Gates’s and Allen’s short-term plans, someone had appropriated an early version of Micro-Soft’s BASIC, stored on paper tape, at a small MITS trade show held in Palo Alto in 1975. The tape was promptly reproduced and then handed out at such venues as the Homebrew Computer Club, a semilegendary group of computer hackers and enthusiasts who met regularly in Silicon Valley to share information, gossip, advice, and other things, such as “liberated” chips and especially liberated Altair software. Soon, paper tapes containing an early, buggy version of Altair BASIC were in wide use, and oddly enough, no one offered to pay Micro-Soft a dime for the product.
In 1975 there was very little that was kumbayah about Bill Gates, and he responded to the purloining of Microsoft BASIC by writing an open letter to the software liberators, published in the Homebrew Computer Club’s newsletter (and in similar publications), chiding them for their thieving ways and asking them to voluntarily pay for the privilege of using his BASIC. His letter made the logical point that if people weren’t recompensed for all their time and hard work spent creating new and better software products, they would have no incentive to do so, and the software industry would wither and die.
Gates’s pleas for financial remuneration went widely unheeded. The very act of releasing the letter generated generous amounts of sneers and opprobrium from software’s kumbayahs, three hundred or four hundred letters addressed to Gates chastising him for his greed, and about three or four voluntary payments for Altair BASIC. Ruined by the premature widespread release of Altair BASIC and the financial loss this entailed, Micro-Soft went out of business, and Gates and Allen were never heard from…aga…errr…no. That’s not what happened.
What actually happened was the widespread release of Altair BASIC established the product as the de facto standard for microcomputers. Despite some idiosyncrasies, Micro-Soft’s BASIC was regarded as an engineering triumph—lean, loaded with features, and, in comparison with the mainframe and mini-computer BASICs most programmers worked with, incredibly fast. Although everyone didn’t want to pay for Altair, which later became Microsoft (with no hyphen) BASIC, everyone wanted to use it. Since Microsoft’s deal allowed the company to license the product to other firms, Microsoft was soon enjoying a tidy business licensing its BASIC to a plethora of other computer companies. In point of fact, it was the industry’s high regard for Microsoft’s BASIC that led IBM to Bill Gates’s door and enabled him to take advantage of the biggest business opportunity of the 20th century.
Nonetheless, as the industry began its rapid development, resentment on the part of software entrepreneurs grew as software piracy spread. And make no mistake, spread it did. Copying a software program worth hundreds, or even thousands, of dollars was as easy as inserting a blank floppy disk into a disk drive and typing in your system’s version of the “copy” command. Games in particular were the target of frequent liberation efforts, with user groups for systems such as the Amiga and Atari ST sponsoring “swap nights” where members were encouraged to bring in their software collections for communal sharing. Many businesses entered into the kumbayah spirit of things, with it being a common occurrence for a company to buy one copy of a business software package such as WordStar and distributing it to every member of the company.
To counter the practice of software liberation, now usually called “piracy,” a whole host of what were eventually called “copy protection” systems and techniques were developed. Most of these focused on protecting Apple software because this computer system attracted the bulk of new software development until the release of the IBM PC. Some of the techniques employed included things such as forcing a disk drive to write to locations on a floppy nominally off limits to the hardware; “Spiradisk,” a system that wrote data to the disk surface in a big spiral; hardware “dongles,” plastic keys that contained a chip with a software key embedded into it; and so on.
In response to the efforts of one part of the software industry to prevent pirating software, another part promptly launched an effort to thwart the protectors (this had the happy effect of employing more programmers). Anticopy protection systems included software products such as Locksmith, copy-cracking boards that sucked an entire software product into memory and spit it out to disk, products that were capable of reading dongle keys, and so on, and so on, and so on. As soon as one copy protection scheme was introduced, it was immediately under attack by resourceful folks following in the glorious tradition of Altair BASIC and the Homebrew Computer Club.
By the early 1980s, IBM entered the market with its own microcomputer, and the focus of the endless cat-and-mouse game between the Capitalists and Kumbayahs shifted to the PC. The software industry’s reaction to rampant software piracy was the general introduction of copy protection for many of the major software packages. WordStar 2000, Lotus 1-2-3, dBase, and other packages incorporated elaborate schemes meant to halt, or at least slow, the piracy tide. For a brief period in the 1980s, almost a dozen software companies were pitching other software companies on the effectiveness of their respective protection systems.
I initially had a great deal of sympathy for the effort. As a field software engineer for MicroPro, I had become quite accustomed to walking into a customer’s location and seeing multiple copies of WordStar (which was not copy protected) installed on every computer in the place but being able to spot only one set of manuals available to the “user” base. Some simple math seemed to indicate a lot of bread was being snatched from my mouth, or at least from the mouth of the company paying my salary.
It was also annoying to find myself spending time providing technical support to people who were clearly flying the software Jolly Roger. One of my responsibilities was to take local technical support calls while in the office from people who were having difficulty with our word processor. A disturbingly high number of my calls went something like this:
Me: Hi! This is MicroPro technical support. How can I help you? The “customer”: I need help installing my NEC 3550 printer.
Me: No problem! Please pull out your installation manual out, and turn to page 256. (This was an age when users were a manly bunch, with thumbs thickly muscled from paging through software documentation similar in size and comprehensiveness to small encyclopedias. Not the like the effete perusers of PDFs and HTML you find today.) I’ll be glad to walk you through the process.
The “customer”: Uh, I don’t have a manual in front of me.
Me: No problem. I’ll hold on the phone until you can get it.
The “customer”: Uh, I don’t have a manual.
Me: Can I ask what happened to it? The “customer”: Uh, the dog ate it. (Other popular claims focused on thieving kids, roaring fires, and torrential flooding.)
The computing press (the members of which were used to obtaining all the free software they wanted) was, as you might imagine, generally unsympathetic to the plight of the software firms. Despite giving perfunctory lip service to the idea that software companies had a right to protect their property from theft, the companies were (and are) constantly being lectured on “not treating their customers” like thieves, despite the indisputable fact that large numbers of them were (and are). In 1984, MicroPro estimated that eight pirated copies of WordStar were in use for every one sold. In 2005, estimates put software piracy rates in China at more than 90 percent.
And yet, by the end of the 1980s, practically every software that had implemented copy protection dropped it. Several factors were driving this trend. One was that many companies resisted buying copyprotected software because it added complexity and instability to desktop computing systems and strained the resources of IT departments. Another was that copy protection added considerably to the software industry’s support burden because users called up to complain about systems that wouldn’t install because of hardware peculiarities, lost or damaged “key” disks, arguments about the number of “valid” installs, and so on. And, although our feelings undoubtedly weren’t the strongest factor driving corporate decisions, most software firms were hearing whines and groans from their field sales and support personnel about the difficulty of dealing with protected products. WordStar 2000, for example, at one time used a copy protection system that limited users to three installations of the software on different systems. This meant that whenever I or another person had to install WordStar 2000 on a demo system at a remote location, we had to go through a wearying install/deinstall routine while listening to outraged disk drives go AAAHHHHKKKK SKRRRIIIKKK WAAKA WAAKA WAAKA in order to keep our quiver full of demo installs for future use. (Field personnel weren’t initially given non-copy-protected products. When we were, the practical facts we created “on the ground” provided another reason to drop copy protection.)
And finally, despite the theoretical losses software companies were suffering from piracy, it was hard to see in reality how piracy was hurting the companies. As the decade progressed, many software companies did indeed stumble and fall, but in no case was it possible to pin the blame on piracy. Also, it started to become apparent to software firms that piracy had a definite upside, as Microsoft had discovered years ago with the Altair. When the number of people using your software increased, your perception as the market leader increased as well. And pirated software functioned as a sort of marketing kudzu, tending to choke out the competition as use of your product spread throughout the computing populace. Once you had displaced the competition, it was possible to convert X percent of the pirates to paid users via various inducements and offers. Corporations, worried about legal liabilities, were also usually not reluctant to buy purloined software if the price was right.
Becoming the market leader also opened up opportunities for bundling and original equipment manufacturing (OEM) deals. At MicroPro, WordStar’s early ubiquity made it the favored word processing product to include with such systems as the Osborne, Kaypro, and many others. While OEM products were sold at a considerable discount from the software’s retail price, in most case all the software publisher had to do was provide licenses and serial numbers to its customers; the OEM customer usually was responsible for manufacturing and supporting the product. One MicroPro OEM salesman referred to the firm’s OEM business as a “money-printing operation.” This model worked in the case of such products as WordStar, dBase, WordPerfect, and most notably, Microsoft Windows. Today, Microsoft’s Windows OEM business is the most profitable component in the company’s bottom line.
In the meantime, while the proprietary software companies were garnering all the attention (and making all the money) from the market, the kumbayah forces, led by an interesting fellow by the name of Richard M. Stallman, were keeping the dream of free software alive. Stallman had entered computing by way of MIT in 1971, where he worked as a systems programmer in the university’s AI lab, at that time a hotbed of innovation in such areas as LISP and related languages. Stallman developed a reputation as an ace programmer and while at MIT developed the legendary program Emacs, a text editor backed up by a powerful and extensible macro system. Stallman was a militant believer in what was then called the “Hacker Ethic,” a belief system that preached that software and the information it represented should be open and available to all users to change and modify as they saw fit. Stallman was fervent in his belief about the evils of charging for software, at one time proclaiming that “the prospect of charging money for software was a crime against humanity.”
Unfortunately for RMS, as his friends called him, by the 1980s the MIT lab was becoming corrupted by the sirens of commerce, who asked why geeks couldn’t also have fancy cars, big homes, and gorgeous girlfriends. Two AI companies (both ultimately unsuccessful) dedicated to building LISP interpreters and dedicated LISP machines spun out of the MIT lab, taking with them many of the lab’s best programmers and all, in the opinion of RMS, of the lab’s kumbayah mojo.
After a period of mourning, Stallman left the lab with a vision fixed firmly in his imagination. He would create a powerful, free, and open software environment that would allow programmers to create new and wondrous products. This environment would be based on the popular (but proprietary) UNIX operating system and, in a display of geek wit, would be called GNU (GNUs not UNIX; I’m sure you appreciate the recursion). And to ensure that what had happened at MIT could never happen again, he’d protect this environment with a new and innovative concept, a “copyleft” agreement that required programmers who used his software to build new software to make the original GNU software, and any changes or improvements made to the software they had created, available for free to anyone who wanted it under the GNU General Public License (GPL). When the GPL was introduced, Stallman became software’s Dr. Open, the civilized, reasonable, humanitarian advocate of all that was good and pure in the world. (Bill Gates has traditionally played the role of Mr. Proprietary, but since he’s supposed to be leaving Microsoft to cure diseases worldwide, Steve Ballmer will be appearing in the part moving forward.)
This was a sharp and revolutionary contrast with the typical enduser license agreement (EULA) that accompanied most proprietary software. Most EULAs allowed “licensees” of software only the right to copy “their” software onto a limited number of computers. In fact, by 2006 the Microsoft retail EULA for Windows allowed you to copy your $100+ copy of Windows XP onto only one computer, regardless of how many computers you owned. And boy, oh boy, better make sure you never, ever, buy a four-core processor in your computer, because that seemed to violate the Microsoft EULA. And if you read the rest of the EULA, it warned of all kinds of other things you couldn’t do, and all the warnings were written in the Scary Lawyer dialect of the English language. In fact, most EULAs are full of scary language and all kinds of implied legal threats. Interestingly enough, despite that software companies have been using EULAs for decades, it is unclear whether they have any legal validity. Fortunately for the industry, no one actually ever reads an EULA; if they did, everyone would probably use only free software....