One of the depressing things about the Massachusetts economy is how many startups here get bought up by firms from other states, particularly California.  Lots of good ideas start here, but move elsewhere before they scale up to significant size, or disappear altogether.  The most egregious recent example is Facebook, which began at Harvard but whose founders were told to move to Palo Alto in order to be close to capital and programming talent.  Other examples are rife: Apollo Computer, which pioneered graphical workstations before being bought by HP in ’89 and dissolved; Unisphere Networks, builder of core routers, bought by Juniper in 2002; and the biggest of all, Digital Equipment, bought by Compaq in 1998 (then HP in 2002) with a big piece bought by Intel in ’97.   Of the three firms where I’ve spent most of my career, two have been bought by CA firms: the semiconductor group of DEC (bought by Intel as mentioned above), and Pixel Magic, which was bought by Oak Technology and then by Zoran, both of CA.

The lamp appears in every interview of this documentary

I recently came across a documentary about an interesting case of this, the text adventure game company Infocom.  The documentary is called “Get Lamp”, and mainly consists of interviews with people involved in this odd sub-genre of computer gaming.   The text adventure genre is now defunct (i.e. is still pursued by obsessive fans), but they were huge in the 80s with titles like “Zork” and “The Hitchhiker’s Guide to the Galaxy”.  I played that one myself and did in fact get the babel fish.  Zork and Visicalc were the reasons to buy a PC in 1980.

Infocom was founded by MIT people, and was remarkably sophisticated in both its software and its gaming style.  In those days there was not a single dominant type of PC, so they built an interpretive engine, the Z-machine, that let the same game run on almost any hardware.  They pioneered an object-oriented programming style with inherited classes because they wanted coders to be able to specify a set of actions that one could do with an object in the game (E.g. “get lamp”, “light lamp”, “throw lamp”) and then propagate them to lots of other objects.  This is all a decade before Java.

The games themselves were also great advances over the earlier Adventure and Colossal Caves games.    The puzzles were good, the descriptions of the settings were evocative, and some scenes even quite moving.   The scenario of the game “A Mind Forever Voyaging” (a reference to Newton) would have made a great SF novel.   “Leather Goddesses of Phobos” had levels of raciness for descriptions: “tame”, “suggestive”, and “lewd”.  The games also had cool packaging: a blank face for “Suspended: a Cryogenic Nightmare”, and a flying saucer for “Starcross”.   “Get Lamp” has its own nice “feelie” – a coin with the iconic lamp and the motto “Where do you want to go today?”

By the mid-80s, though, the graphical computer games were killing the text adventures.   These all came from California, unsurprisingly.  Infocom resisted the trend, believing that the player’s imagination would supply better visions than that of the crude displays of the day.   Images versus words is an old California versus Northeast struggle, and images win.  A CA firm, Activision, who was the leading maker of cartridge games, bought Infocom in 1986.  Relations between them soon soured, and Infocom closed up only three years later.  Activision itself ran into trouble soon after, and nearly went bankrupt.   They saved themselves by packaging up all the Infocom games into one big release, “The Lost Treasures of Infocom” 1991, and sold $10M of them, enough to restart.  They’re now one of the largest game operations in the world.

That must have been aggravating to the Infocom people, but I wonder if it really mattered.   Most of them went on to other good work.   Their VP of Marketing, Mike Dornbrook, is now COO at Harmonix, maker of “Rock Band”.  One founder, Marc Blank, later wrote the “Syphon Filter” series and an email reader for the Palm.   Another, Bruce Daniels, worked on the Lisa and on Mac database software.   The least satisfying outcome was for the writer of their most sophisticated games, Steve Meretsky, who seems to have mainly done consulting work since.

In the interviews they talk about what a great time they had at Infocom, but a lot of that sounds like it just came from being young and loose and smart.   Even though their work didn’t turn into lasting fame and fortune, they made a real mark at an early stage of their careers, and that’s more than most people get to do.

The same can be said of most purchased MA companies.  The people involved keep with it or go on.  With some exceptions, the technologies persist if they’re worthwhile, or decay if they aren’t.   Companies are only shells, after all.   If one fails you go to the next.  That’s hard to accept when you’re in the middle of a failure, as I’ve been, but you need to take the longer view.  What matters are the people involved and the work they do, not the name on the stock certificates, or the US state of the owners.

[PS A detailed description of the fall of Infocom can be found in a report prepared for an MIT class in 2000 on the Structure of Scientific Revolutions:  “Down From the Top of Its Game – the Story of Infocom”.  They made some other key mistakes, but it’s hard to succeed when your product is being superseded.]