RE: FOR070.DAT files appearing
- From: "O'Brien Paddy" <Paddy.O'Brien@xxxxxxxxxxxxxxxx>
- Date: Mon, 13 Feb 2006 19:44:24 +1100
Hmm, I'll hand insert chevrons, so a lot of avoided comments
-----Original Message-----
From: Ken Fairfield [mailto:my.full.name@xxxxxxxxx]
Sent: Saturday, February 04, 2006 11:22 AM
To: Info-VAX@xxxxxxxxxxxx
Subject: Re: FOR070.DAT files appearing
Bill Gunshannon wrote:
And some of us really don't much care what ISO does!! Driven mostly
by international politics it is a safe bet that ISO will choose whatever
is most likely to annoy the US.
My god... This makes the words, "head in the sand", pop into
my mind...but I won't go there. It does seem highly out of character
coming from someone in an academic environment where international
collaboration, exchange of ideas and cooperation on projects is the
norm. ;-(
Hmm, and Bill is an avid Fortran programmer :-) I'd use the term parochial in the wider sense of country.
I'd also challenge you to find a commercial Fortran 77 compiler
still being sold today. Oh, you can buy a compiler that will compile
Fortran 77 code, but most (by far) commercial compilers available
today, on any platform, are F95 compilers...with a few F90 compilers
as stragglers. Even on the open source front, you'll find g95 and
gfortran, both F95 compilers, but g77 has not been touched in years...
(One exception I will grant is the VAX Fortran compiler which is F77
plus extensions, a lot of them. It does cost money to license and so
counts as "commercial". But VAX was EOL a long time ago, and NO
development has been done on that compiler since the VMS/Alpha
version was released, so it's stretching the idea to say that its
"being sold today".)
This brings to mind my latest concern. Fortran was stopped at F77 on VAX, and the impression I get is that it will be stpped at F95 (with the TR) on Alpha -- whether any further development is done on IA64, who knows. It is essentially in maintenance mode and there seem to be only a handful of egineers to correct serious bugs. As you may remember, Ken, we were both beta testers on all the Alpha versions of F90/F95 from its first appearance. The last beta testing sort of stopped dead when Compaq/HP gave the Fortran team to Intel. As beta testers, we had excellent follow up from Steve Lionel with obviously good backing from his team. But so much has been left in the air. No cross referencing (which has some advantages over SCA), array notation not fully/very optimised (try timing COMPLEX MATMUL against F77 style loops), constants not unfolded -- just to name a few of the deficiences, several of which I reported in the last beta, but which dissolved with Steve's "passing".
When I went to the technical update in Sydney last year, I was given the impression (also because of a bug report I filed) that CXML may never make it to IA64. The reason, no interest. Since it is bundled with Fortran, how is this known. Definitely all my applications use the BLAS and LAPACK from there. Yes, I can re-compile from Netlib or revert to NAG, but I have lost the heavy optimising that was put in and which made one of our eigenvalue applications run in 25% of its earlier time using the NAG versions of the same routines. I am not putting NAG down, they have provided a service by giving us these routines with the original names as entries, but they are not machine optimised in the same way. And, of course, they provide many other excellent routines that we use mainly for integration and optimisation.
Regards, Paddy
***********************************************************************
Please consider the environment before printing this email.
"This electronic message and any attachments may contain privileged
and confidential information intended only for the use of the
addressees named above. If you are not the intended recipient of
this email, please delete the message and any attachment and advise
the sender. You are hereby notified that any use, dissemination,
distribution, reproduction of this email is prohibited.
If you have received the email in error, please notify TransGrid
immediately. Any views expressed in this email are those of the
individual sender except where the sender expressly and with
authority states them to be the views of TransGrid. TransGrid uses
virus-scanning software but excludes any liability for viruses
contained in any attachment.
Please note the email address for TransGrid personnel is now
firstname.lastname@xxxxxxxxxxxxxxxx"
***********************************************************************
.
- Prev by Date: Re: null terminated strings
- Next by Date: Bind server strategy in cluster
- Previous by thread: Re: FOR070.DAT files appearing
- Next by thread: Boot Error on Shadowed Sys Disk.
- Index(es):
Relevant Pages
|
