Relative measures

Wed 10 Sep 2003, 04:30 PM

Tweet

by Ben Langhinrichs
I look forward to Tuesdays, because on Tuesdays, the New York Times has a section called Science Times.  Science Times is always interesting, and this last Tuesday was no exception.  Amir Aczel wrote an article about celestial measurements which advocates changing from A.U., light-years, parsecs and megaparsecs to something he calls jet-years.  At roughly 600 miles an hour, a jet travels 5 1/4 million miles a year, which is a jet-year.  This is somehow easier to understand than a light-year.

In a similar vein, Ned Batchelder writes about lines of code per month.  This makes slightly less sense to me than a parsec (about 3.26 light-years).  First, I haven't the foggiest clue how many lines of code I write per month.  Second, I cannot begin to imagine why it would be a useful measurement.  We need measurements that are relative to what matters, and that are easier to understand.  Tell me how many gripes-per-week an employee voices, and I'll tell you whether that employee is destined for the front office or the front stoop.  Tell me how many meeting-shortening-ideas-per-month, and I'll tell you whether this he/she has the right stuff.  In direct programmer terms, tell me how many useless-code-cycles-per-week the programmer eliminated, how many non-obvious-comment-lines were generated, and how many previously-written-subroutines-reused-per-day the programmer has, and I'll know whether this person is who I want stranded on a desolate consulting gig with me.

What measurements would you like to know about a developer/programmer/consultant?  Don't worry about how impossible they would be to measure.  Think outside the box.

Copyright © 2003 Genii Software Ltd.