From mboxrd@z Thu Jan 1 00:00:00 1970 X-Spam-Checker-Version: SpamAssassin 3.4.5-pre1 (2020-06-20) on ip-172-31-74-118.ec2.internal X-Spam-Level: X-Spam-Status: No, score=-0.5 required=3.0 tests=BAYES_05 autolearn=ham autolearn_force=no version=3.4.5-pre1 Date: 5 Aug 93 15:41:57 GMT From: world!srctran@uunet.uu.net (Gregory Aharonian) Subject: Re: Fortran Interface in Ada9X RM - Annex M. Message-ID: List-Id: >If anyone out there knows of an Ada83 compiler company that's going after >all those Fortran folks by implementing this functionality, please let >me know, because I have a LOT of engineer colleagues who might respond >favorably. Their present response about Ada is "Oh, that DoD toy. >They're just a bunch of real-timers who don't give a damn about the >engineering world. Let me know when they'll give me a finite-element >system, or when I can hook up all that old Fortran code, or IMSL." Given that until quite recently, more code was reused in Fortran by scientists and engineers than any other language (now surpassed by reuse of C and C++) code, there are tremendous amounts of Fortran code heavily used around the world, code capability which they will want to have in any new language before they switch. So Mike's concerns are very valid. Unfortunately, given the crimes in coding you can commit with Fortran (and I like the language), much of this is hard to interface to with any pragma convention, especially when it comes to array manipulation. Fortran makes assumptions about array allocation that are not overly compatible with Ada (or any other language). The best solution is to reengineer the Fortran code into Adatran, having studied this problem in great length (since most of my government source code database is in Fortran). For example, consider one of the classic, heavily reused DoD Fortran programs, LOWTRAN. This would have been a very influential code to convert to Ada as a start of converting the hordes of DoD Fortran programmers to Ada. And at 20,000 lines of Fortran, 10,000 of which are mostly arrays of coefficients, it would not require much funding to convert the thing to Ada, if anyone really cared about the problem of DoD Fortran research vis-a-vis the Mandate. But they don't, it hasn't, and Mike's colleagues' excuses remain. -- ************************************************************************** Greg Aharonian Source Translation & Optimization P.O. Box 404, Belmont, MA 02178