[
Lists Home |
Date Index |
Thread Index
]
At 06:51 PM 5/7/2002 +0100, Bill de hÓra wrote:
> > From: Jonathan Robie
> >
> > But the fact that there are errors you can't catch until
> > runtime does not
> > imply that you shouldn't catch other errors before running a
> > program.
>
>Contrariwise, the fact that there exist errors you can catch at
>compile time does not imply they should be caught. Trading machine
>efficiency for programmer efficiency is a spectacularly bad idea
>given what we know about the bottlenecks in developing software
>systems.
I always thought that catching errors at compile time optimized both
programmer efficiency and machine efficiency.
It optimizes machine efficiency because the query doesn't have to look at
all that data to know that there's an error. It optimizes programmer time
because the programmer does not need to go out and write a million test
cases, deal with the calls from the field from someone whose data looks
different from the test data, etc.
Jonathan
|
