What every compiler writer should know about programmers [pdf]

Re:What every compiler writer should know about programmers ... by j2kun on Thursday March 03, 2016 @07:38PM (#11221155)

I half expected the document to present statistical evidence about how programmers tend to behave, and use that to generate advice for compiler writers.

The article is a nice read (and doesn't present that kind of statistical evidence), but I'm still wondering, has anyone done a significant study of how programmers work? I would be really fascinated to read an analysis of, say, data taken from full-time software engineers at a given company where all of their actions are recorded (key-strokes, compiler runs, error messages, etc.). Similar to how online competitive gamers have their actions recorded to the microsecond to identify weak points.

It would even be interesting to know, e.g., what is the net average number of lines of code created/deleted per day?

[ Parent | on HN ]

Re:What every compiler writer should know about programmers ... by conceit on Thursday March 03, 2016 @07:44PM (#11221198)

The average doesn't matter, specific problems are reported on the mailinglists, for example. You wouldn't sway everyone's writing style with some statistical evidence, not in a giant ecosystem like C/unix and it's not a democracy either where 51% would just ignore the other 49% existence.

[ Parent | on HN ]

Re:What every compiler writer should know about programmers ... by j2kun on Saturday March 05, 2016 @12:48PM (#11230162)

What? No, I mean that's how to allocate your resources well, you prioritize the problems that will affect most of the people most of the time.

[ Parent | on HN ]