The Quiet Bullpen

There is something innately wrong when the room is quiet and everyone is engaged in solo work. It's okay for email, or chat, or writing reports (I suppose) but it doesn't seem like software work to me. I suppose that's because software development is a more passionate, collaborative effort. I expect to hear a lot of conversation, arguments, laughter, frustrated groans or exclamations. I expect to hear one person telling the other that they've just forgotten a semicolon or spelled a variable name wrong, or at least that there is quicker way to make some modification.

Software development should be like a jam session, where all the musicians are trying trading between leads and accompaniment and finding spaces to fill in and learning to join in the song without getting in its way. It's noisy in these intellectual software jam sessions, and much gets done. Even those on the sidelines hear the process of software and are able to join in or sit out as their needs and. skills dictate. It is an event of controlled chaos, a feast of intellect and technique.
But no. In the quiet bullpen there are a number of people "minding their own business". There are no offers of corrections, no arguments, no exchange of ideas or techniques. It's dead calm. It gives me the sense that there is something happening but that something is not software design. It's too sterile.

Oh, yeah, there are lines of code being written, but I fear that too many lines are being written relative to the amount of work being done, since nobody is sitting over the typist's shoulder reminding him that it could be done in fewer words, in fewer keystrokes, in fewer clicks. I fear that the same procedure is being performed multiple different ways and nobody realizes it. I fear that the same problems are being corrected in multiple bodies of code. Quiet isolation breeds waste.
I fear that nobody is doing "corrective steering." I fear that any work that comes out of the quiet bullpen will make a quick round-trip through the testing department (if there is one, and woe unto the solo programmer if there is not) back to the programmer. I fear even more that the same programmer will work on it the second time. Silent isolation is a breeding ground for mistakes.

I fear that there are great ideas being generated, but that these ideas are falling into code that is "owned" by one programmer, and her coworkers may never see it. In the silence, nobody can hear your epiphany.
I fear that local optimizations are compromising the larger application and nobody knows it.
I fear that someone thinks his work is done because he forgot about some aspect, some task that should have been considered (and may have been discussed already) but nobody is riding shotgun on his memory.

In the worst case, I fear that people working solo can say or believe anything that they want about their work, and nobody is there to keep them honest. Isolation breeds arrogance, and hides dishonesty.

I trust that people are able to think better in groups. I trust that people are natural problem solvers and that they can pool their natural genius and individual perspective. I think that having many advisors increases success. I think that we're all smarter together than any of us alone.

But in the quiet bullpen, they choose (intentionally or accidentally) not to harness the power of other people. Instead, we get the best effort of isolated individuals with their varying levels of skill, taste, and experience. The quiet bullpen is where mistakes are born, and allowed to breed.

---

!commentForm