Re: ::scr Extreme Programming's Pair Programming

[Greg McCarroll <greg@xxxxxxxxxxxxxxxxxxxxx>]

* Richard Clamp (richardc@xxxxxxxxxxxxx) wrote:
> So are you saying that a pair would be comprised of two snot-nosed
> kids, or one newbie and a seasoned hacker, or something different
> altogether?
> 

i have no idea, i was hoping some of the members of the church
of XPology could offer suggestions, however in their absense lets
look at the combinations,

	newbie + newbie:
		definetly an improvement on newbie on its own,
		the assumption being that where one doesnt know
		how to do something, the other might know.
		i'd guestimate that you'd get 1.25*newbie productivity
		at 2*newbie cost (BAD VALUE)

	newbie + seasoned:
		in reality the seasoned programmer will occupy the
		keyboard 75% of the time (if not its probably even
		less productive), this will end up with a 
		0.25*newbie + 0.75*seasoned result, and some training
		by product, however the training by product will
		be incredibly poor compared to GOOD training courses
		(BAD VALUE)
	
	seasoned + seasoned:
		they will argue over interesting problems and probably
		quite enjoy the banter, but at best this is gonna give
		you 1*seasoned as value (BAD VALUE)

so imho, its all bad.

> Since you only want to talk pair programming lets put the "proper
> software development processes" handwave into a different thread,
> though I am interested to hear what you'd consider proper and improper
> software development processes to be.

ok start the thread with a statement of opinion and i'll buy a ticket
for the ride.

-- 
Greg McCarroll                                 http://217.34.97.146/~gem/