On 2005-07-01 08:14-0700 Walt Brainerd wrote:
> Alan W. Irwin wrote:
>> Walt, that is an excellent suggestion from Arjen. The INSTALL file
>> documents how to compile and link all the examples and execute them.
>> Furthermore, the INSTALL file documents how to compile and link your own
> Thank you both again for the suggestions. I tell my
> tale below, not just to show my stupidity, but to
> maybe give you some ideas when fixing up the docs
> that might save others some grief.
> Late last night I figured it out. But ...
> I did run the examples. That convinced me that things
> were working. The Makefile has variables nested in such
> ways that it looked very difficult to actually see what
> is needed to compile.
> I also looked at the stuff in Chapter 21 of the PDF doc.
> When I ran plplot-config with --libs, I guess I could
> not absorb the idea that I needed that many options.
> Rereading it says exactly that, however. Also, the
> with-f77 option described in INSTALL is not mentioned.
> Later I stumbled on the better instructions in INSTALL
> and those got things working. If just those few lines
> were added to Chapter 21, I think that would be a big
> help. It was dumb not to find the Cygwin installation
> instructions in INSTALL, but I was not expecting
> compilation instructions to be there.
> Anyway, I now need to learn more about how PLplot
> actually works. That should be more interesting and
> a lot more fun!
> As with the other stuff I put with the Fortran Tools,
> I will try to make it a little easier for a newbie
> to get going. Since I will provide the binaries, it
> should be easier.
> PLplot looks like a great plotting tool. Those who
> built it should be proud. Thank you.
> One last thought. g95 seems pretty solid and the
> other GNU Fortran compiler gfortran probably will
> be some day soon, so the 95 version of Fortran
> should be included. If, some day you do that, I
> would be willing to modify all the Fortran examples
> to be written using "modern" Fortran 95. That would
> be my contribution for using your software and
> help. Of course, Arjen and others could do it
> easily, but obviously they have other things to
> do to contribute to PLplot.
Thanks, Walt, for your comments on the documentation. I find a fresh eye
(like yours) sees documentation problems much better than veteran users of
I have had many years of fortran experience, but I stopped paying attention
to fortran advances at about the Vax/VMS era because I didn't have research
money to pay for commercial compilers. Later on, when I switched from Unix
to Linux, g77 never gave me any problems with my Vax/VMS-era research
fortran programmes, but didn't support modern fortran so again there was no
motivation to learn anything about modern fortran. However, now that g95 is
finally getting pretty solid it is on my agenda to learn modern fortran, and
I expect that is a decision other free fortran users are beginning to make
as well. Thus, in this context of free fortran change, it would make a lot
of sense to provide a modern fortran interface and examples for PLplot
(although for backwards compatibility we would want to retain the current
fortran 77 interface and examples indefinitely as well).
The problem as always is time/energy for developers to provide a modern
fortran interface and examples for PLplot. Walt, if you are interested in
working on such a project, I would be happy to test your patches and apply
them to our cvs version. Note, although we have discussed using cfortran.h,
we currently have a home-brew method of providing our current f77 interface.
To avoid name clashes between the routine names used by external users and
our own internal routines, we provide names for our common API that start
with "c_". See plplot.h. The common API is that subset of our public C API
that every interface (including fortran) to PLplot should implement. The
actual fortran and C routines we use to implement the current fortran 77
interface are in bindings/f77.
Alan W. Irwin
Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).
Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); PLplot scientific plotting software
package (plplot.org); the Yorick front-end to PLplot (yplot.sf.net); the
Loads of Linux Links project (loll.sf.net); and the Linux Brochure Project