Related articles |
---|
Fortran to Java translator cwong1@cs.uiuc.edu (David Wong) (1998-12-19) |
Re: Fortran to Java translator abate@brahma.ticam.utexas.edu (1998-12-21) |
Re: Fortran to Java translator ttlegrad@plg2.math.uwaterloo.ca (1998-12-21) |
Re: Fortran to Java translator wclodius@aol.com (1999-01-03) |
From: | ttlegrad@plg2.math.uwaterloo.ca (Tom Legrady) |
Newsgroups: | comp.compilers.tools.javacc,comp.compilers |
Date: | 21 Dec 1998 11:51:12 -0500 |
Organization: | University of Waterloo |
References: | 98-12-055 |
Keywords: | Fortran, Java, arithmetic |
David Wong <cwong1@cs.uiuc.edu> wrote:
>I am going to write a Fortran 90 to Java translator and I would like
>to use JavaCC to generate the parser.
W. Kahan & Joseph Darcy point out a number of reasons why Java produces
incorrect results in intensive mathematical calculations;
see
http://www.cs.berkeley.edu/~wkahan/JAVAhurt.pdf
or
http://www.cs.berkeley.edu/~darcy/JAVAhurt.pdf
Java mathematics is quite satisfactory for most applications, and
is easier to grasp than a number of other languages, thus it is fine
for ordinary applications.
But in applications where precise and correct results to complex
computations are required, Java is unsatisfactory. Unfortunately, it is
exactly those applications which are likely to be written in Fortran.
Perhaps you should consider converting Fortran to some other language for
compilation.
Tom Legrady
--
ttlegrady@plg.uwaterloo.ca (519) 888-4567 ext 4822
DC 3548 - the PLG Lab home: 725-6003
[I really don't want to get the floating point thread from last month
started again, please consult the archives if you want to see lots of
previous comments. -John]
Return to the
comp.compilers page.
Search the
comp.compilers archives again.