Wed, 31 May 2000 13:54:40 +0400 (MSD), S.D.Mechveliani <[EMAIL PROTECTED]> pisze:
> in quotRem, quot, rem,
> divMod, div, mod.
>
> the latter triplet is unnecessary, if the first has the mode argument.
Mode is unnecessary, better to have separate functions here.
> And we can hardly invent the mode type better than Char,
> because any specially introduced mode types bring the long names.
>
> quotRem 'n' x (-3) looks better than the pair quotRem & divMod,
It does not. And since you say you like short names: it's shorter.
I cannot imagine a case where this parameter would not be constant.
In 99% of cases either I know which I want or it does not matter.
Also
map (`quot` 2)
looks better for me than
map (\n -> quot 'n' n 2)
Char is bad because Char means a character. Far from a choice of
definition of modulus. It would be a misleading message to the reader.
Types are not only sets of representable bit patterns; they also carry
a meaning. Otherwise Char, Int and Bool would all be one type. In
C they almost are, and that is one of reasons I use Haskell rather
than C.
GHC sometimes uses types like
data CanItFail = CanFail | CantFail
instead of Bool. Maybe even too rarely: when somebody is writing
a piece of code, he remembers what True or False or -1 mean in a
particular place, but for another person who reads it, it is not
that obvious. Haskell uses LT,EQ,GT instead of an integer <0,==0,>0
(even OCaml uses <0,==0,>0). GHC's Posix module has a separate type
Fd for file descriptor instead of using Ints. Types are good.
--
__("< Marcin Kowalczyk * [EMAIL PROTECTED] http://qrczak.ids.net.pl/
\__/ GCS/M d- s+:-- a23 C+++$ UL++>++++$ P+++ L++>++++$ E-
^^ W++ N+++ o? K? w(---) O? M- V? PS-- PE++ Y? PGP+ t
QRCZAK 5? X- R tv-- b+>++ DI D- G+ e>++++ h! r--%>++ y-
