In 1995, I was one of Jim Gray‘s first hires for his new research lab in San Francisco. Jim didn’t enjoy managing, so he never let the lab grow beyond about a dozen lab members. Consequently, I spent the following ten years enjoying regular small-group interaction with a “database legend” (as he was called on the cover of Database Magazine).
Jim won the Turing Award for “seminal contributions to database and transaction processing research and technical leadership in system implementation” and tech luminaries lined up at his packed-out memorial service to give him credit for his impact on the modern world of transaction processing: “every time you use an ATM machine you can thank Jim Gray.”
But Jim was very self-deprecating about his fame. “I just wrote it down,” he told me.
Ted Codd, of course, was the one who came up with relational algebra, and Jim minimized his own role as merely writing a tech report about building one of the first relational database systems (System R). This report ended up being read and cited by nearly everyone in the industry and Jim ended up famous – almost by chance, it seems, in his account. (Never mind that his genius gave us ideas like ACID, the five-minute-rule, and the data cube)
So the lesson he drilled into me was: “It doesn’t exist unless it’s written down.”
I would tell him my project was finished, and he’d reply: “No, it isn’t. Where’s the report? Where’s the memo? Where’s the web page?” He wanted something that others could find, that he could point to as the completed output. If it wasn’t written down, it wasn’t done. It didn’t exist for anyone to refer to, so it just didn’t exist. He also used to insist on a written monthly report, listing what I had accomplished the previous month and what I planned to do in the coming month.
Over time, I came to appreciate the wisdom of what Jim said. I internalized it to the point that I continued writing monthly reports even for managers who didn’t ask for them, and I always made sure that what I did ended up written down.
Now this is a value for my team at Trōv. In the life of a distributed organization it has more meaning than ever. We aren’t in the same time zone, never mind in the same hallway, so it is more important than ever that things are written down. “Where’s the wiki? Where’s the shared doc?” – “It doesn’t exist unless it is written down”
In part two of this post, I’ll speak more about this rule at Trōv, and also describe its potential dark side.
Originally posted at the Trov Engineering Blog