Mailing Lists: Apple Mailing Lists

Image of Mac OS face in stamp
 
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Bundling Fortran WAS: Re: [apple scitech] installation of LAM/MPI



On Nov 16, 2007, at 3:19 PM, Ernest Prabhakar wrote:

As David Gohara can tell you, this is something we've been discussing with MacResearch for several years. The challenge has been waiting for a stable version of gFortran for gcc 4, since g77 wasn't drop-in compatible with gcc 3.



Since Ernie brought this up, I think we are OK to start discussing this more openly. It is correct that we've been pushing for this for some time. A few months back we put up a request to help come up with a test suite to test gfortran builds:


http://www.macresearch.org/macresearch_request_fortran_numerical_test_suite

We didn't give much reason for this at the time since we weren't allowed to publicly discuss this until now.

The point that Apple has made is that it wants to support fortran development, but there are some issues in doing so. The primary being that Mac OS X doesn't have a single line of fortran in it. If they break gcc they know immediately since they build Mac OS X every night and problems can be addressed immediately. They also have a vested interest in gcc since, again, it is used to build the OS.

What that means is that we need to come up with a good testing suite that covers the gamut of compilation scenarios and can also be used to test the numerical accuracy of output from the builds. There are some testing facilities that come with gfortran, but in reality given the variety of code and applications that use fortran, we need something that tests the tool and tests against other fortran code bases.

What I would propose is that as a user community of scientists and developers we come up with a set of reliable tests that Apple could then use for QA for building (and including) gfortran as part of the developer tools. Ideally we'd keep this focused in one or two places (max), so either on this list or as part of the thread at the URL above. If there are also any compiler developers (either gcc/gfortran or otherwise) that can also lend guidance or are willing to participate, we'd welcome any contributions and insight into things that should be considered.

I personally am glad that Ernie commented here. In the past people have voiced their opinions about the lack of fortran support in the Apple dev tools, this is our chance and we should collectively move on this. Otherwise there is nobody to blame but ourselves.

  Regards,

Dave

David W. Gohara, Ph.D.
Center for Computational Biology
Washington University School of Medicine
http://gohara.wustl.edu
http://www.macresearch.org
314-362-1583 (phone)
617-216-8616 (cell)




_______________________________________________ Do not post admin requests to the list. They will be ignored. Scitech mailing list (email@hidden) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/scitech/email@hidden

This email sent to email@hidden
References: 
 >Bundling Fortran WAS: Re: [apple scitech] installation of LAM/MPI (From: Ernest Prabhakar <email@hidden>)



Visit the Apple Store online or at retail locations.
1-800-MY-APPLE

Contact Apple | Terms of Use | Privacy Policy

Copyright © 2007 Apple Inc. All rights reserved.