SunSoft: The B Team

Right about the time I joined Windows Engineering at Sun, the SDF was introduced.

Sun had recently released it's first version of Solaris (as opposed to SunOS) and things hadn't gone particularly well. It was big, buggy, and slow. Rob Gingell explained this (paraphrasing), "We hired a couple of thousand engineers and guess what? They all wrote code!"

So Rob had taken a detour from being a God in engineering to lead a team to set up some way of herding the cats to deliver a Product Life Cycle type of process, and that ended up being the Software Development Framework. This probably cost Rob a couple of points on the deity scale with the engineers, but he could afford to drop a few for the benefit of figuring out how to deliver Solaris.

This was not fun for organizations that had been used to calling their own shots and the initial steps were daunting and frustrating. Actually, that never stopped but we got used to it and the processes continued to evolve.

One of the characteristics of the SDF was that everything was organized by Teams. W Team, C Team, E Team. Actually I can't even remember what all the teams were at this point, but it did seem that every letter of the alphabet had been co-opted into service of the SDF.

But one letter did remain.

Milind Pansare stuck his head into my office one day, probably in the fall of 1992 or spring of 1993. He announced that I should come to a B Team meeting right away in the break room.

I happily trotted along and found that a handful of the engineers had gone out and bought 2 or 3 six packs. B stood for Beer and that promptly became the code word for our own mini beer bust (the big official ones having been recently banned).

Before long, more people joined in and B conveniently became Beverage as not everyone drank alcohol.

Sooner or later, this reached the ears of Real Management (not just the likes of me). Our Senior Director and then VP, Paula Sager was a good sport. B Team became a monthly event for the entire group. We all had to contribute a penny for the beer to prove that we weren't coerced into drinking. It lost a bit of the spontaneity and evil thrill but Paula did pick up the tab and provided alternative options like ice cream sundaes. This continued for a long time, even after I left for IT almost 6 years later and Milind and so many others had pushed on as well.

Once in those early days of grappling with the SDF and getting used to the group, I attended a team meeting led by Steve MacCay, our VP at the time.

Steve had the unenviable task of addressing the growing (or contraction) pains of SDF and making us enthusiastic about it.

Steve was a typical introverted nerd. I think he had gotten a Ph.D in something like studying the asexual reproduction of bacteria.

Nevertheless, Steve did have some oil in the crank case and he did deliver a couple of very pithy observations from time to time, many of which I remember. In this case, he said that when he was asked why he was working on this Solaris stuff and his answer was, "Because it's hard."

He then went on to explain that some things simply do not scale as they grow. Software engineering is one of those. What works when you're a small, nimble startup just doesn't hang together with thousands of engineers delivering code. There were no easy answers and that was one of the intriguing things.

Steven didn't make this concept up, of course, and I have no idea how enthusiastically he believed this, but this lecture made a big impression on me and helped guide my thinking about a long list of projects and environments I've had to deal with over the years.