PC-Talk

~ the eulogy ~

I missed this one by about sixteen years and an ocean. PC-Talk was written in late 1982, when I was still an infant in Turkey and a personal computer was a thing almost nobody in the country had seen. By the time I got online in 1998 on a 14.4k modem, the program itself was long gone and the idea it started was so completely absorbed into how software worked that nobody attached a name to it anymore.

That idea is why this grave exists. Every nag screen I ever clicked through and every honor-system program off a shareware CD glued to the front of a Turkish computer magazine descends from one man's decision in 1982 to put a program on a floppy disk and ask nicely.

Andrew Fluegelman was not a programmer. He was a lawyer who had quit law, then a book publisher running a one-man operation called The Headlands Press out of Tiburon, California. In October 1981 he bought one of the first IBM PCs sold in San Francisco, having agreed to co-author a book called Writing in the Computer Age. Within two weeks he was writing his own accounting software in BASIC. He later said he spent the first month programming and loved it. The next year he needed to send book drafts to a coauthor over a phone line. Nothing on the market did it the way he wanted, so he wrote PC-Talk instead.

~ the dossier ~

~ the pledge drive ~

He did not copy-protect it. His own explanation was that he had been programming for a matter of months and did not know how. Friends told him to sell it, and instead he built a distribution scheme out of the only fundraising model he happened to admire, which was the pledge drive on KQED, the public television station across the bay.

The mechanism was this. Send Fluegelman a formatted floppy disk and he would mail it back with PC-Talk on it. Use the program, and if you liked it, send $25. If you did not, do not. Either way, make copies for your friends, and he shipped a batch file to make the copying easier. He called the arrangement Freeware and registered the word as a trademark, though the term that stuck to the model was one somebody else coined later. At the time he also described it as user-supported software, which is a better name for what it actually was.

He was clear that this was not charity. He called it an experiment in economics more than altruism, and the experiment returned a result nobody had predicted. By 1984 he estimated that as many as half of PC-Talk's users were paying him. The checks came in fast enough that he described the response as literally driving him out of business and had to hire two employees to keep up with the mail. He also noticed a second effect: because the software was not sold, users sent him suggestions and bug fixes instead of angry reviews.

~ what it started ~

He was not alone for long, and he did not try to be. Jim Knopf, a Seattle programmer with a database called Easy-File, was doing something similar at the same moment. The two of them coordinated deliberately. Knopf renamed his program PC-File to match and set it at the same suggested price, and each man agreed to mention the other's product inside his own documentation. Two men who had never met arranged a shared brand for a distribution model that did not have a name yet.

PC-Talk's own community did the rest without being asked. The source was available, derivative versions multiplied, and the CompuServe IBM/PC forum produced a maintained build labeled PC-TALK III Version B, Level 850311, a build string that reads as March 1985 and appears to be the last officially sanctioned release of the program. Members of the Houston user group HAL-PC made their own variants with videotex and IBM 3101 terminal emulation. Fluegelman and The Headlands Press, who held the copyright, blessed all of it. Larry Magid reviewed the program in PC Magazine and said it was elegantly written and performed beautifully.

~ july 1985 ~

By 1985 Fluegelman had become a magazine editor. He had launched PC World in 1982 and Macworld in 1984 and was editing both, which meant PC-Talk got less of him every year while other terminal programs took the ground it had held. Procomm was already on its way.

He had ulcerative colitis. In July 1985 he was put on prednisone for it, and within days of starting the drug he became agitated and depressed, apologizing repeatedly to colleagues for failures they had not noticed. His employer rearranged his schedule to take pressure off him. It did not help. On the afternoon of 6 July 1985 he left his office in Tiburon. A week later his car was found abandoned at the north end of the Golden Gate Bridge. His body was never recovered and his family held a memorial service. He was forty-one.

No source records a date when PC-Talk was discontinued, because it never really was. There was no announcement to make. The copyright sat with a small publishing house whose sole proprietor was gone, the last sanctioned build was already four months old, and the program simply stopped being updated while the model it invented went on to underwrite an entire decade of software. This grave is dated to the day its author disappeared, because that is the only honest date available for it.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »