scs@adam.pika.mit.edu (Steve Summit) (04/29/89)
In article <4402@goofy.megatest.UUCP> djones@megatest.UUCP (Dave Jones) writes: >Somewhere or another I heard of the "telephone test" for programs. >I'm sorry I can't credit the author. I haven't seen the answer mentioned, so if anyone is still curious, the idea appears in Kernighan and Plaugher's Elements of Programming Style. I agree it seems like an extreme test, but my opinion of that book is that it's so deeply correct that any disagreement is probably due to imperfect sophistication on the part of the reader, so I'm not prepared to say I won't be wholeheartedly advocating the telephone test at some later date. I _h_a_v_e had the experience, many times, while trying to explicate some code of mine to someone else, in a code review or something, of realizing that what I had thought was a justifiable, neat trick was actually incomprehensible and indefensible. Generally it's best to err on the side of simplicity and readability. Steve Summit scs@adam.pika.mit.edu