Re: Intrinsic functions in fortran 90
- From: glen herrmannsfeldt <gah@xxxxxxxxxxxxxxxx>
- Date: Mon, 11 Apr 2005 03:31:34 -0700
Jan Vorbrüggen wrote:
It seems that early FORTRAN did not have mixed mode expressions at all. There must have been a lot of XINTF function calls. I do hope you don't mean to go back that far.
Indeed, I would. It would certainly be a useful feature in a compiler to provide warning of implicit conversions - I believe some do.
In my mind, this is similar to the question of typing of literals.
Well, Java requires casts for narrowing conversions but not widening ones. That doesn't seem so bad to me. I believe it will do int to float without complaining, but not float to int. That doesn't solve the 5/9 problem, but might get a good fraction of the other mistakes.
-- glen
.
- References:
- Intrinsic functions in fortran 90
- From: student
- Re: Intrinsic functions in fortran 90
- From: Bob Walton
- Re: Intrinsic functions in fortran 90
- From: student
- Re: Intrinsic functions in fortran 90
- From: Gary L. Scott
- Re: Intrinsic functions in fortran 90
- From: student
- Re: Intrinsic functions in fortran 90
- From: Richard Edgar
- Re: Intrinsic functions in fortran 90
- From: glen herrmannsfeldt
- Re: Intrinsic functions in fortran 90
- From: Richard E Maine
- Re: Intrinsic functions in fortran 90
- From: glen herrmannsfeldt
- Re: Intrinsic functions in fortran 90
- From: Jan Vorbrüggen
- Intrinsic functions in fortran 90
- Prev by Date: Re: Is it ok to ask about C++ code / C++ compiler problems on comp.lang.fortran ?
- Next by Date: Re: Intrinsic functions in fortran 90
- Previous by thread: Re: Intrinsic functions in fortran 90
- Next by thread: Re: Intrinsic functions in fortran 90
- Index(es):
Relevant Pages
|