DSNES(Y,N,FUN,EPS,IT,AS,BS,MS,IER) Fortran 77 sources wanted !
la...@rts.ua
One old man whose business of life is develop of machinery for mines,
lost the subj. It is from IMSL 77 .
Becasue he lost it, he is not able to recover very important model
written in Fortran IV for mainframes long time ago . He has it at some
tape, but it is no chance to find the working tape device now.
He has already this routine for older Fortran version, like Fortran90,
but he is too old to adapt it to the old Fortran IV model.
I want to help and to make him happy.
So, if some has this source, in any form, even printed one, please send
it.
Regards,
Igor from Ukraine
dpb
la...@rts.ua wrote:
> Hi, Dear ALL !
>
> One old man whose business of life is develop of machinery for mines,
> lost the subj. It is from IMSL 77 .
>
> So, if some has this source, in any form, even printed one, please send
> it.
...
IMSL routines are only distributed as object libraries to link to user
code. Afraid you'll either have to find a compiler which includes the
IMSL license and object files (the Professional version of the
DEC/Compaq Visual Fortran, for example, altho it is no longer
supported/sold) or purchase the IMSL license directly from Visual
Numerics or another source, or replace the functionality of the missing
routine with another routine.
Actually, when I look, I don't find a routine of the name DSNES in the
IMSL documentation so don't know it would help, anyway. Are you
absolutely sure this was from IMSL library?
Also, regarding the version of Fortran. Later versions essentially are
supersets of earlier standards so for computational code such as this,
the difficulty in compiling an existing earlier piece of code should
not be excessive with whatever compiler one is currently using.
Don't know if that gets you anywhere, but the end result is it doesn't
appear to me the routine as listed in the subject is actually IMSL,
anyway even if source were available. Actually, I just did a search of
the VisualNumerics online documentation for the F77 libraries for
"DSNES" and found no hits there, either...
la...@rts.ua
Thank you very much for reply.
Yes, it is from IMSL 77 - he is absolutely sure. However it could be
Russian version of IMSL77 from Russian clones of some IBM mainframe (ES
series, ES 1045 for instance).
It is listed in more than 20 years old russian reference books .
This program solves the system of transcendental equations.
As for newer version of DSNES - he has it. I have seen on his computer
the file DSNES.F90 - something like this. ( I am not Fortran guru, I
wrote my last Fortran operator also 20 years ago).
And I can confirm - it was not obj, but Fortran source file. And it is
not suitable because it require several external (?) missing modules.
I suppose in 1980s the sources of IMSL 77 was available in USSR...
However the object file will help as well.
I also tried to find it at VisualNumerics and fails.
It seems the problem is too specific to time/territory...
I afraid he will be forced to rewrite it from the scratch....
Anyway, thank you !
dpb писал(а):
beli...@aol.com
> Hi,
>
> Thank you very much for reply.
>
> Yes, it is from IMSL 77 - he is absolutely sure. However it could be
> Russian version of IMSL77 from Russian clones of some IBM mainframe (ES
> series, ES 1045 for instance).
>
> It is listed in more than 20 years old russian reference books .
>
> This program solves the system of transcendental equations.
>
> As for newer version of DSNES - he has it. I have seen on his computer
> the file DSNES.F90 - something like this. ( I am not Fortran guru, I
> wrote my last Fortran operator also 20 years ago).
>
> And I can confirm - it was not obj, but Fortran source file. And it is
> not suitable because it require several external (?) missing modules.
>
> I suppose in 1980s the sources of IMSL 77 was available in USSR...
>
> However the object file will help as well.
>
> I also tried to find it at VisualNumerics and fails.
>
> It seems the problem is too specific to time/territory...
>
> I afraid he will be forced to rewrite it from the scratch....
There exist public domain Fortran codes to solve systems of nonlinear
equations, for example
http://users.bigpond.net.au/amiller/NSWC/hbrd.f90 . I suggest that he
try to write his own version of dsnes which just calls one of these
codes. The information needed is what the arguments of DSNES mean.
Richard Maine
> IMSL routines are only distributed as object libraries to link to user
> code.
While that is true today, it wasn't always so. IMSL used to distribute
source code, and with a very liberal license, at that. You were allowed
to freely redistribute the source code of the procedures that your
application needed. I know that seems surprising in today's environment,
but I very distinctly recall reading the license for the copy of IMSL
that we had; it was pretty clear (i.e. in English instead of legalese).
I also recall their license going rather strongly in the other direction
when they were bought/merged/whatever with, um, Visual Numerics I think
it was. The Visual numerics license was so strict that it didn't
actually allow us to use what we had paid for. :-( We had a development
license for one machine, and a site-license for running the executables
that had IMSL-linked routines in it. Yes, we had to pay extra for the
license to run our executables on other machines. But it turned out that
you could not run the executables without making copies of some of the
support files that the license said we weren't allowed to copy. I recall
being disgusted that the very liberal older IMSL license was replaced by
one so strict as to make it literally unusable. That was a while ago;
things have probably changed again since then,
So I find it quite believable that the OP might once have had IMSL
source code, particularly if his application dates from FORTRAN IV days.
I don't have any direct help for him, I'm afraid, but I thought I'd note
that I find his story plausible.
--
Richard Maine | Good judgement comes from experience;
email: last name at domain . net | experience comes from bad judgement.
domain: summertriangle | -- Mark Twain
Michael Metcalf
"Richard Maine" <nos...@see.signature> wrote in message
news:1hlav8c.kb13utsaordN%nos...@see.signature...
>
> So I find it quite believable that the OP might once have had IMSL
> source code, particularly if his application dates from FORTRAN IV days.
> I don't have any direct help for him, I'm afraid, but I thought I'd note
> that I find his story plausible.
>
much energy to copying/stealing hardware, software and complete systems from
Western sources. I would bet a large sum of money that the source code used
by OP's friend was never legally acquired and paid for!
Regards,
Mike Metcalf
dpb
Richard Maine wrote:
> dpb <dpbo...@swko.net> wrote:
>
> > IMSL routines are only distributed as object libraries to link to user
> > code.
>
> While that is true today, it wasn't always so. IMSL used to distribute
> source code, and with a very liberal license, at that. You were allowed
> to freely redistribute the source code of the procedures that your
> application needed. I know that seems surprising in today's environment,
> but I very distinctly recall reading the license for the copy of IMSL
> that we had; it was pretty clear (i.e. in English instead of legalese).
>
Hmmm....the only place I worked that had IMSL on mainframes was in the
late 60's and thru '78 and we were certainly not given source code,
only linkable libraries at that time. What they would have done w/ a
government organization I don't know. The requirements at that
location were pretty onerous in keeping even the documentation under
control of a special set of "knowledge dispensers" but we could
redistribute the object files of any support routines required w/ the
application.
After leaving that organization I went w/ a very diversified consulting
firm that didn't buy anything centrally so until licenses were made
available for the PC at a price-point I was able to sneak it into a
contract of my own I never had access again until the bundling w/ CVF
which I sprung for myself.
So, if you saw it work that way I'm not going to say otherwise, but I
never saw it operate that way, but granted there was a long gap from
roughly '78 to approximately '95 I wasn't in an environment that
allowed me any access at all. I've not had to deal w/ VisualNumerics
directly.
glen herrmannsfeldt
(snip)
> Actually, when I look, I don't find a routine of the name DSNES in the
> IMSL documentation so don't know it would help, anyway. Are you
> absolutely sure this was from IMSL library?
I have the 1984 version of the IMSL Users manual, and it isn't there,
either. It looks like both Fortran 66 and Fortran 77 routines
are described. A google search also doesn't find it.
Chapter D is Differential Equations and Quadrature (integration),
I don't know if that is what he means by transcendental equations.
-- glen
Fly Away
Although what you are saying may be true in this particular case, I do
not think you are in position to make such a statement. It does sound
impolite and somewhat offensive (to me). What happened to the famous
"political correctness" and presumption of innocence?
P.S.
Also, please do note that the so called "stealing" mostly took place
for certain military purposes whereas regular scientists/engineers had
nothing to do with it. And of course there was no internet at that
time, which eliminates any possibility of "stealing" on the personal
level.
Michael Metcalf
"Fly Away" <victor....@gmail.com> wrote in message
news:1157731453....@i42g2000cwa.googlegroups.com...
> Also, please do note that the so called "stealing" mostly took place
> for certain military purposes whereas regular scientists/engineers had
> nothing to do with it. And of course there was no internet at that
> time, which eliminates any possibility of "stealing" on the personal
> level.
>
was, alas, systemic at the time.
Regards,
Mike Metcalf
jon
Fly Away wrote:
> Michael Metcalf wrote:
> > "Richard Maine" <nos...@see.signature> wrote in message
> > news:1hlav8c.kb13utsaordN%nos...@see.signature...
> > >
> > > So I find it quite believable that the OP might once have had IMSL
> > > source code, particularly if his application dates from FORTRAN IV days.
> > > I don't have any direct help for him, I'm afraid, but I thought I'd note
> > > that I find his story plausible.
> > >
During the late '70s and the '80s I worked at a place that had IMSL to
use with our minicomputers. We had full access to all the source code.
It was distributed as part of the package we licensed. We didn't even
consider this strange at the time. So it is probable the source code
was available.
Jon