[Haskell-cafe] Re: Sugestion for a Haskell mascot
claus.reinke at talk21.com
Thu Mar 12 09:40:36 EDT 2009
>I agree that looking for a mascot that is inspired by "laziness" is a bad idea
>from a P.R. perspective (I am tired of people walking out the room when I
>give Haskell talks to general audiences and explain lazy evaluation).
Do they walk out when you mention it or when you explain it?-)
Lazy evaluation -as an optimization of non-strict evaluation- seems
like an unambiguosly good idea. Lazy evaluation -as one efficient
form of non-strict evaluation- has its pros and cons: neither strict
nor non-strict evaluation fit all purposes, and the real trick is to
find the right middle ground. It just so happens that non-strict is
a safe default from which to try and reach that middle ground. In
other words, even in non-PR terms, laziness is a stepping stone,
not the ultimate goal.
Your remark reminded me of some old slides of mine, where I
tried to offer one perspective on the problems of "communicating
fp ideas to general audiences". In brief, successful communication
assumes some shared background, and if that doesn't exist,
communication is difficult at best and usually fails.
Haskellers often resort to formal maths models, which is fine for
those with a shared background, not so fine for general audiences.
In that old talk I suggested using a model that general audiences,
and business folks in particular, are familiar with, and started to
outline an initial "dictionary of fp terms" - the translation worked
well enough to show that neither strict nor non-strict evaluation
make for good business models, and that we're really looking for
some form of "just in time" evaluation (of course, you have to keep
in mind that my understanding of business terms is only that of an
average general audience;-).
I've temporarily put the slides here (note that the contact info,
from one of my previous lives, is years out of date):
Perhaps you find some of the ideas useful? And now that we
actually have some more business folks amongst us, perhaps
some of them would like to comment on the suitability or
otherwise of these ideas?-).
-- Lazy evaluation: been there, done that.
More information about the Haskell-Cafe