COMPUTING SCIENCE
Rumours and Errours
Brian Hayes
Gnothi Seauton
Mistakes bring the gift of self-knowledge—a gift that is not
always welcome. Looking back on this episode, I could summarize it
as follows: I wrote a program that gave a wrong answer, and then I
fiddled and fudged until I finally got the output I wanted, and then
I stopped. This is not a protocol to be recommended. What's most
troubling is the uncomfortable thought that if the textbook answer
had not been given to me at the outset, I would surely have been
content with the result of my first, fallacious, program.
Still, for most of us, the only way we'll never err is if we never
try. My fellow-columnist Henry Petroski has written eloquently about
the necessary role of error and failure in all worthy undertakings;
as he says, falling down is part of growing up. And if we are going
to make mistakes, it seems salutary to bring them out in the open
and discuss their causes. Staring them in the face makes them seem a
little less mortifying.
Only a little, though. A confession of this kind is not followed by
absolution. And instead of "Go and err no more," Knuth
quotes Piet Hein's advice: "Err and err and err again but less
and less and less." I take my own motto from the novelist and
playwright Samuel Beckett: "Fail again. Fail better."
© Brian Hayes
» Post Comment