Hi,
>>>>> "Bjorn" == Bjorn Lisper <[EMAIL PROTECTED]> writes:
Bjorn> I imagine expressions like x==x could appear as the result of
Bjorn> transformations in program specialisation, like specialising
Bjorn> f x x
Bjorn> where
Bjorn> f x y = if x==y then ....
Bjorn> Replacing x==x with True is of course still unsafe in
Bjorn> Haskell, for the reasons already pointed out.
that's true, of course. Furthermore, aside from "x==x", optimizations
could deal with "x-x", "x && (not x)" and many other expressions.
Concerning "safety". As far as I can see there is only a single reason:
Partially correct programs terminate more often. Doesn't
that meet the philosophy behind laziness?
Christoph
- RE: How to force evaluation entirely? Ch. A. Herrmann
- RE: How to force evaluation entirely? Patrik Jansson
- Re: How to force evaluation entirely? Lennart Augustsson
- Re: How to force evaluation entirely? Fergus Henderson
- RE: How to force evaluation entirely? John Hughes
- Re: How to force evaluation entirely? Marcin 'Qrczak' Kowalczyk
- Re: How to force evaluation entirely? Bjorn Lisper
- Re: How to force evaluation entirely? Carl R. Witty
- Re: How to force evaluation entirely? Michael Marte
- Re: How to force evaluation entirely? Zhanyong Wan
- Re: How to force evaluation entirely? Ch. A. Herrmann
- Re: How to force evaluation entirely? Ch. A. Herrmann
- Re: How to force evaluation entirely? Carl R. Witty
