Why I owe Adam Stoller three apologies

In the late 1980s and early ’90s, I worked on the Andrew Project at the Information Technology Center at Carnegie Mellon University, where Adam Stoller was a fellow programmer and system administrator. On at least three occasions I opposed Adam’s advocacy of something or other, only to reverse my opinion later.

Adam kept trying to interest me in XPilot, a space-battle game; but I found it confusing and could never get interested in it. Years later I became a huge XPilot fan. I’ve even contributed several new features.


American pop culture regarded sushi askance during my formative years.

In a 1978 episode of Columbo, “Murder Under Glass” [directed by Jonathan Demme!], chef Louis Jourdan serves a special sushi meal to a visiting Japanese dignitary, who exclaims, “Fugu sashimi — in Los Angeles!”

In The Breakfast Club (1985) there’s this exchange between Judd Nelson and Molly Ringwald: “What’s that?” “Sushi.” “Sushi?” “Rice, raw fish, and seaweed.” “You won’t accept a guy’s tongue in your mouth, and you’re going to eat that?”

And from Desperately Seeking Susan that same year: “Now you have these sushi restaurants. Everyone goes for sushi.” “Sushi… I hate the stuff.” “Although, I tell you, I had some the other day. I took it home, I cooked it, it wasn’t bad. It tasted like fish.”

Adam was also known for his abiding love of sushi. At that time, sushi hadn’t permeated American culture to the extent that it has now. I found the mere idea of eating raw fish to be repellent, and never hesitated to say so to Adam — perhaps even cruelly. Now hardly a day goes by that I don’t crave a meal of sushi.

Once, after a considerable research effort, Adam tried to convince the ITC to switch from using RCS for source code control to using CVS. CVS was fairly new at the time, and I was among those who were distrustful of CVS’s fundamentally different approach than that of RCS. It didn’t require files to be locked before being edited; instead, if two developers edited the same file at the same time, they were later obliged to use a dubious merging algorithm to resolve any conflicts that may have arisen. Collectively we defeated Adam’s proposal and continued using RCS. Within a few years, though, CVS became the de facto standard for source code control in organizations with multiple software developers; no one even considered using RCS anymore. I’ve even contributed some bugfixes to CVS.

Adam, you were right about so many things, and I was wrong.

Leave a Reply