Lan Barnes wrote:
On Tue, March 18, 2008 11:09 am, mark wolfe wrote:
[EMAIL PROTECTED] wrote:
On Mon, Mar 17, 2008 at 07:07:40PM -0700, Andrew Lentvorski wrote:
People seem to be finally shrugging off the Extreme Programming(TM!) as
unsubstantiated BS.
At my old job there were 2 programmers that went through school together
that always programmed everything together. They were taught pair
programming
as The Way to program.
Man, one of them will be screwed if the other gets hit by a bus. :)
Mark
Is it just me, or do most of the "new" paradigms seem to enshrine the
undisciplined, undergraduate, "let's have fun" mentality?
Well, pair programming is probably one of the better suggestions in the
book, IMO. However, it almost certainly doesn't offset the productivity
loss. For me, pair programming has advantages, but productivity isn't
one of them. For example, actually having to *read* code for a change,
actually having to explain yourself, watching people use their tools,
moving knowledge from the experienced to the novice--these are all
useful tasks.
Of course, you could argue that good groups manage to do this anyway.
That's probably true, but the process looks a lot like pair programming
when you actually examine it in the end. Good groups have a lot of
"Okay, let me pull up a chair and take a look at what your doing..."
Acutally documenting a thought process before you start to code isn't
_that_ bad, is it?
It depends. Sometimes yes, sometimes no. If I'm having to be very
exploratory, documentation will be as useless as unit tests. It's a
question of malleability vs. benefit.
-a
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-lpsg
