Lists Home |
Date Index |
Dare Obasanjo wrote,
> It's always amusing to see technologists blindly reciting anecdotal
> statements about software development as if they are the laws of
> Physics or were handed down from above on a fiery mountain.
Except that Postel's is a better candidate for law of nature status than
At the risk of stating the blindingly obvious: "be liberal in what you
accept and conservative in what you send" can be read as an analog of
the "contravariant argument types, covariant return types" function/
method typing rule familiar from the polymorphic typed lambda calculus
and manifested in at least a few modern programming languages with well
designed type systems.
Given the solid basis for the rule, I'd say that there's a presumption
in it's favour other things being equal. Other things aren't always
equal tho', so there are going to be good grounds for local and limited
violations. But in those cases the burden of proof is on the violators
to come up with a persuasive justification.
That's a bit weaker than a law of nature, but it's a good bit stronger
than blind acceptance of anecdote.