All the Perl that's Practical to Extract and Report

XP pre-history

inkdroid on 2004-01-28T22:48:55

I can't remember if I saw this article cited here on use.perl, but I printed it out months ago and finally got around to reading it. I had no idea that there was such a rich history of iterative development that stretches back to the 1950s...way before XP was a glimmer in Kent Beck's eye. The piece kind of reminded me of what Damian said once when visiting chicago.pm about how much of XP is just good common sense, and how it really isn't "extreme" at all :-)

XP Prehistory

dws on 2004-01-29T00:18:43

A number of XP's practices are quite old--Pair Programming dates back to punchard days, when the benefit of having an extra set of eyes look for problems could save a one day (or worse) compile/test cycle--but the genius of XP is in bundling the practices together. The self-supporting combination becomes much stronger than any one (or two or three) practices done in isolation.

Re:XP Prehistory

ziggy on 2004-01-29T00:48:52

I think it's a bit of a stretch to date Pair Programming back to punched cards. There was a lot of 'desk checking', and asking others to look over card stacks and printouts. And, with the high cost of computing, there was a lot of debugging at the desk (both before and after a run), so many times, the second set of eyes was for debugging, not programming per se.

But some of that is just picking nits. It doesn't matter a whole helluva lot if this is the earliest instance of pair programming, or a prototype for pair programming.

The earliest instance I remember coming across is in a book somewhere[*]. The author interviews a consultant from the early days of TTYs, where he had one TTY in the living room (probably a dialup line to a mainframe where they rented time). They would program in pairs in front of the TTY because they simply didn't have enough money or room to have a terminal for each programmer, so they wrote code in pairs.

 

*: Probably in one of the XP books, something vaguely about agile programming, or something recent by Tom Demarco or Gerry Weinberg. (Sorry, that's as close as I can remember off the top of my head.)

Re:XP Prehistory

dws on 2004-01-29T01:04:32

I think it's a bit of a stretch to date Pair Programming back to punched cards.

Jerry Weinberg told me that John von Neumann's team at IBM used Pair Programming in much the same form that XP employs it now. That'd be back in the days of punchcards. And Jerry is savvy enough to understand the difference between Pair Programming and desk checking decks.

Re:XP Prehistory

ziggy on 2004-01-29T01:13:32

Fair enough.

It doesn't jive with my memories or experiences before I was using a dedicated terminal/machine, but I wasn't programming with John von Neumann or Jerry Weinberg. ;-)