>>[On KISS] I use this principal sometimes but try to avoid it and
always look past the obvious forgetting about any existing barriers in
your way
Just exactly so. I find it used this way too, most often.
And yet, I rework software almost beyond measure trying to
simplify it, and after a while, by gum, even the most
difficult problems usually end up with a simpler solution
(algorithmically sometimes, but most often just implementation).
To put it another way, which I think is really what you're
getting at, simplicity takes a hell of a lot of work.
On the other hand, if you're approaching something you're
just beginning to understand, keeping it simple and direct
upfront saves you from a lot of useless work which will be
thrown away after some experience, when you understand it
better. I think this is the sense in which Paul is using
the term, despite his unfortunate reference to software
design. It has taken me many, many years to learn this
(ok, I'm a slow learner), and I STILL err on the side of
analyzing too early, before I know enough to do an
efficient analysis. Throw most of it out and could have
done it better with a few more mistakes under my belt.
KISS works for this. Sad to say, the distinction is very
difficult to make. Of course, this is called a "top down"
approach in the jargon, but it's not simple to use
the discipline effectively.
To broach an old subject a bit, since I've laid
down all this groundwork <g>, I find a VERY great deal of
popular project management, software "quality" methodology,
and similar hot software topics very much in this category.
I specifically include structured walkthroughs, code reviews,
etc. I have seen them work, BTW, but far (far far) more
often, they are reviews of ill-understood problems by
people ill-prepared to judge what they are reviewing
based on the ill-considered notion you can anticipate
end results with a "careful" design. May I never ever
hear the term "careful design" again from anyone who
can dictate my activities ...
Sorry, an over-lengthy aside on an ancient topic.
Spots |
