trh@ukc.UUCP (06/14/84)
Somebody was asking for info on this but I've lost the address so hope this may be of general interest. There is a special issue of Nasti News (Oct 1983) by Pat Gaffney, Oak Ridge National Laboratory, P.O. Box Y, Oak Ridge, Tennessee 37830. He asks the question "Is it feasible to program an IBM PC in Fortran to do scientific computations?" His conclusion is that as a result of minimal testing of the Microsoft FORTRAN compiler, version 3.1 it is possible. But it is only tolerable with a hard disc and an 8087 coprocessor. This configuration will allow "fairly modest scientific programming to be performed relatively easily". The new compiler from Microsoft is " a significant improvement over the one distributed by IBM", although it only implements a subset of f77. Finally "The time periods required for the COMPILE-LINK procedures will probably necessitate a change of work style for most programmers. I have found that these periods are sufficiently long to be useful for eating, drinking and sometimes even thinking !" Tim Hopkins, Computing Laboratory, University of Kent, Canterbury CT2 7NF Kent U.K. { trh@ukc.UUCP }