Post #135,573
1/13/04 10:34:24 PM
|
OpenWatcom 1.2 released
[link|http://www.openwatcom.org|http://www.openwatcom.org]
--\n-------------------------------------------------------------------\n* Jack Troughton jake at consultron.ca *\n* [link|http://consultron.ca|http://consultron.ca] [link|irc://irc.ecomstation.ca|irc://irc.ecomstation.ca] *\n* Kingston Ontario Canada [link|news://news.consultron.ca|news://news.consultron.ca] *\n-------------------------------------------------------------------
|
Post #135,600
1/14/04 3:46:19 AM
|
WOW NICE!
I'm ordering the CD!
Real FORTRAN! Woo-hoo! LINPACK goodness!
-drl
|
Post #135,604
1/14/04 3:59:22 AM
|
ICLRPD (new thread)
Created as new thread #135603 titled [link|/forums/render/content/show?contentid=135603|ICLRPD]
Peter [link|http://www.debian.org|Shill For Hire] [link|http://www.kuro5hin.org|There is no K5 Cabal] [link|http://guildenstern.dyndns.org|Home Page - Now with added Zing!]
|
Post #135,642
1/14/04 9:16:57 AM
8/21/07 5:53:01 AM
|
You don't like g77?
Part of the gnu compiler suite? Has support on Mac OS X BTW.
"I believe that many of the systems we build today in Java would be better built in Smalltalk and Gemstone."
-- Martin Fowler, JAOO 2003
|
Post #135,651
1/14/04 9:49:39 AM
|
It's fine but it's not "real FORTRAN"
It's a front end to a C compiler. A real FORTRAN compiler can do amazing optimizations because it has less to worry about in terms of memory juggling and linking.
Intel/Linux compiler benchmarks: [link|http://www.polyhedron.co.uk/compare/linux/f77bench_p3.html|http://www.polyhedro.../f77bench_p3.html] Best overall, Lahey ("your FORTRAN source since 1967!")
Intel/Win32 compiler benchmarks: [link|http://www.polyhedron.co.uk/compare/win32/f77bench_p3.html|http://www.polyhedro.../f77bench_p3.html] Best overall, Intel
I assume these benchmarks are mostly on "real FORTRAN 77" code, without dynamic allocation etc.
-drl
|
Post #135,696
1/14/04 12:41:58 PM
|
On that note
The scitech guys are funding a feasibility study to port openwatcom to linux.
--\n-------------------------------------------------------------------\n* Jack Troughton jake at consultron.ca *\n* [link|http://consultron.ca|http://consultron.ca] [link|irc://irc.ecomstation.ca|irc://irc.ecomstation.ca] *\n* Kingston Ontario Canada [link|news://news.consultron.ca|news://news.consultron.ca] *\n-------------------------------------------------------------------
|
Post #135,823
1/14/04 10:28:10 PM
|
Confused here.
Unless you have source for huge Fortran programs that you cannot practically port to say C.
Probably before everyone else's time, there was a language called [link|http://wombat.doc.ic.ac.uk/foldoc/foldoc.cgi?Michigan+Algorithm+Decoder|MAD] (Michigan Algorithm Decoder) and a compiler to go with it. One of the associated programs was MADTRAN that translated (i.e. automatically ported) FORTRAN programs to MAD. MAD programs ran much faster.
The last time I wrote in Fortran, IIRC, was in 1972. Fortran should be deader than COBOL! :)
Alex
The mind is not a vessel to be filled but a fire to be kindled. -- Plutarch
|
Post #135,839
1/15/04 1:18:10 AM
|
Re: Confused here.
I have the occasional need to actually crunch numbers, as in running a gravity code - these are always dependent on well-known FORTRAN libraries. Despite heroic efforts, C and C++ have not replaced FORTRAN as the language of choice for numerical computing.
-drl
|
Post #135,844
1/15/04 2:35:59 AM
8/21/07 5:56:56 AM
|
I think many people have moved to matlab for that
At least for explorational crunching of a casual/developmental nature.
"I believe that many of the systems we build today in Java would be better built in Smalltalk and Gemstone."
-- Martin Fowler, JAOO 2003
|
Post #135,847
1/15/04 3:09:16 AM
|
Fizzy Cysts still use FORTRAN
At least according to Matt Nobes (manobes on here, fizzy cyst by trade).
Peter [link|http://www.debian.org|Shill For Hire] [link|http://www.kuro5hin.org|There is no K5 Cabal] [link|http://guildenstern.dyndns.org|Home Page - Now with added Zing!]
|