PLplot

"Alan W. Irwin" wrote:
> 
> Hi Arjen:
> 
> Just to follow up on this topic, there has been no response to my bug report
> either from Michael Lemke (the guy who originally noted the problem several
> years ago) or the Cygwin mailing list.  So it appears everybody is content
> not to deal with this g77 packaging/linking issue on Cygwin even though it
> should be straightforward to sort it out with the nice test results you put
> together for me.  It is possible we are victims of g77 developer fatigue with
> gfortran and g95 taking up more and more free fortran compiler mind share.
> 

Pity. It was to be expected, but one would expect at least an answer.

> Anyhow, for the combination of g77, Cygwin, and shared libraries I have
> modified the cvs HEAD version of f77.ac to disable command-line parsing.
> 
> I have confirmed that the change does not interfere with the correct Linux
> result (to allow fortran command-line parsing), but my cf/*.ac skills are a
> bit shaky (I just copy syntax and logic from one file to the next, and it
> usually works) so could you please check this gives the correct result (a
> bindings/f77/configurable.f that just returns with an error message rather
> than attempting to do command-line parsing) for the combination of g77 and
> shared libraries on Cygwin?  Also, please check that the fortran part of
> plplot-test.sh now works properly for shared-library Cygwin systems.
> 

I should have Cygwin properly set up now as a compolete development 
environment (including CVS and such). So that ought to speed up this
kind
of tests. I will report the results ASAP.

Regards,

Arjen

Hello,

I just wanted to inform you about some advances I have made on the
Windows platform:

Using the win32 driver and Compaq Visual Fortran I have been able to
build the FORTRAN 77 bindings for "native" Windows. At least example 1
is now working ;).

I had to make a few adjustments to the code in bindings/f77, due to the
need for a special keyword on the Windows platform (__stdcall). This
also
revealed a few mistakes in the file sfstubs.f that do not show up on
other platforms (string1 used instead of s1 in plgdev() and plgfnam())

I need to consolidate the whole procedure and then commit the stuff,
but as far as I know this has been the first time for this combination
of language and platform.

Regards,

Arjen

Arjen Markus writes:
 > I just wanted to inform you about some advances I have made on the
 > Windows platform:
 > 
 > Using the win32 driver and Compaq Visual Fortran I have been able to
 > build the FORTRAN 77 bindings for "native" Windows. At least example 1
 > is now working ;).
 > [...]
 > I need to consolidate the whole procedure and then commit the stuff,
 > but as far as I know this has been the first time for this combination
 > of language and platform.

Ah, the memories.  You remind me that in the spring of 1991, I was sitting on
the French Riviera (in Nice, our apartment was actually on the waterfront),
using PLplot with MS Fortran on OS/2 (having recently switched away from
regular Windows 3.whatever it was back then).

That was the experience that drove me to switch to C++ (anyone remember OS/2
gcc/emx?) and then ultimately adopt Linux at the ".95c+" release level,
sometime in the early '90's.  X11 hadn't even been ported to LInux yet, I ran
in console mode for some months until it became possible to install X11.
That was back with "jump table" libs, long before ELF and .so's appeared on
Linux. 

My how the world has changed.

Arjen,

I have a completely working plplot using Compaq Visual Fortran with the 
exception of two of the examples (I think 14 was one of them).  From your 
comment about stdcall you probably have most of the changes already but 
you are welcome to my files if you want them.

Bryan Peterson
bryan_peterson@...

On Wed, 7 Dec 2005, Arjen Markus wrote:

> Hello,
>
> I just wanted to inform you about some advances I have made on the
> Windows platform:
>
> Using the win32 driver and Compaq Visual Fortran I have been able to
> build the FORTRAN 77 bindings for "native" Windows. At least example 1
> is now working ;).
>
> I had to make a few adjustments to the code in bindings/f77, due to the
> need for a special keyword on the Windows platform (__stdcall). This
> also
> revealed a few mistakes in the file sfstubs.f that do not show up on
> other platforms (string1 used instead of s1 in plgdev() and plgfnam())
>
> I need to consolidate the whole procedure and then commit the stuff,
> but as far as I know this has been the first time for this combination
> of language and platform.
>
> Regards,
>
> Arjen
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click
> _______________________________________________
> Plplot-devel mailing list
> Plplot-devel@...
> https://lists.sourceforge.net/lists/listinfo/plplot-devel
>

Bryan Peterson wrote:
> 
> Arjen,
> 
> I have a completely working plplot using Compaq Visual Fortran with the
> exception of two of the examples (I think 14 was one of them).  From your
> comment about stdcall you probably have most of the changes already but
> you are welcome to my files if you want them.
> 

Hello Bryan,

I would surely like to compare them. The main problem is, as far as I am
concerned, to get the building process all smoothed out, so that people 
can use it via a simple make command (or equivalent).

Regards,

Arjen