General Magic

» see generalmagic.com as it lived, on the wayback machine

~ the obit ~

general magic sat in the part of computer history i can read about but cannot personally remember. 1990 to 2002. my window into computers started in 1998 with a 14.4k modem in turkey, and the magic cap world had effectively already ended for the consumer market by then. the 2018 documentary is honestly how most people my age learned the name existed.

marc porat had the pitch first. he wrote a memo at apple in '89 about a pocket-sized communicator that would carry your personal data, send messages, find things for you on a network. john sculley read it, agreed it was important, then decided apple could not ship two pda projects at once and spun the team out as general magic in may 1990. the other pda project was the newton. that one stayed inside.

the founding team was bill atkinson, the guy who wrote macpaint and hypercard. andy hertzfeld, who had written most of the original mac toolbox. susan kare did the icons. joanna hoffman was on the business side. the resumes were apple royalty, which made the partner sign-ups straightforward. sony agreed to build the device. motorola agreed to build a wireless variant. at&t, sony, philips, matsushita, mitsubishi all signed on as backers. that part read more like a consortium than a startup, and that turned out to be the structural problem.

the product shipped in '94. sony magic link, eight hundred dollars. motorola envoy, fifteen hundred dollars on the wireless side. the os was magic cap, a desktop metaphor on steroids where your inbox sat in a literal room and you walked down a literal hallway to the postal service. telescript was the agent language that was supposed to let you send out a little program that would book a flight or sort your mail without you watching it. ambitious technology. zero buyers. the device was slow, the screen was small, the wireless network it depended on was at&t personalink which was a pager-tier service. apple's newton came out a year earlier with most of the same problems and absorbed most of the press.

~ the rap sheet ~

~ how it landed ~

the public-facing story is that the consortium model killed it. you cannot run a startup with five fortune-100 partners sitting on your board, each demanding their own market-specific features. honestly, the magic link had to ship with a generic keyboard layout because no single partner could decide which carrier's billing scheme should be the default. the schedule slipped two years. by '94 the netscape browser was about to ship and the network everyone wanted was the open internet, not at&t's walled-garden pager service. magic cap built for the wrong network.

the company tried to pivot in '97 to a phone-based voice service called portico. it kept the company alive for two more years on a different product. by '99 the stock had collapsed from a february '95 ipo peak, with layoffs and missed projections eating what was left of the headcount. on september 18, '02 the board announced the dissolution. paul allen's company bought the patent stack in 2004 for scrap value.

~ what got buried with it ~

the alumni list is the part of general magic that did matter. tony fadell went to apple, built the ipod, then the iphone hardware. andy rubin went to danger, then founded android. pierre omidyar left to start ebay. megan smith went to google then became the us cto under obama. joanna hoffman went back to apple. kevin lynch later ran the apple watch software side. the documentary frames the company as a school that built the people who built every consumer device of the next twenty years, which is a tidy narrative but probably also true.

the device pitch itself, a pocket communicator with your data on it that talked to a network for you, that pitch was correct. it was just eight years early on the hardware and ten years early on the wireless infrastructure. the iphone shipped in '07 with most of the magic cap pitch translated into a touchscreen and a 3g radio. the agent layer that telescript described is the kind of thing people are trying to build again in 2025 with llms. nobody at general magic gets paid in 2026 for being right too soon.

~ the next chapter ~

the 2018 documentary, general magic, is what most people watch first now. it premiered at tribeca, sarah kerruish and matt maude directed, with interviews from porat, atkinson, hertzfeld, fadell. archival video from inside the office in the early '90s. you see a lot of people who would later be famous arguing about whether a magic cap room should have a window. it is a useful artifact for anyone who wants to look at what visionary product work looks like when the market and the technology underneath are both wrong, but everything above the wrong layer is correct.

@}-->--

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

« previous  ·  random  ·  next »