Do One Thing and Do It Well?

January 4, 2016

I came across an old Slashdot interview with Rob Pike today, and was amused to see that he raised some unusually pragmatic ideas on the old Unix philosophy of “do one thing and do it well”. The related question asked by a reader:

Given the nature of current operating systems and applications, do you think the idea of “one tool doing one job well” has been abandoned? If so, do you think a return to this model would help bring some innovation back to software development?

And Pike’s response:

Those days are dead and gone and the eulogy was delivered by Perl.

This is often one of those places where developers suddenly get philosophic to the point where otherwise good programs are written so that they become unnecessarily difficult to use in the name of correctness. The best example of this is still the modern command line; very functional for people like yours truly who enjoy the throwback to the 70s and all the esoteric arcana that comes with it, but which presents an artificially high bar to entry for anyone who hasn’t already memorized the usage of about 50 common Unix programs, 500 associated flags, and how to compose them to do what you want.

A modernized version of the philosophy might run something like “identify core competencies and do them well”. Such a tenet should serve any developer well as long as they keep the converse problem dictated by Zawinski’s Law in mind as a counterbalance (i.e. “every program attempts to expand until it can read mail”).

Do One Thing and Do It Well? was published on January 4, 2016.

Find me on Twitter at @brandur.

Did I make a mistake? Please consider sending a pull request.