On 2016-01-01 14:29-0800 Alan W. Irwin wrote:
> Just to give you an update, that report now reads like this for the
> new Fortran binding project:
>
> f95
> Missing examples : 16 19 20 21 22
> Differing graphical output :
> Missing stdout :
> Differing stdout :
>
> with perfect valgrind results for all non-missing examples as well.
That report has now been changed to
f95
Missing examples : 19
Differing graphical output : 16
Missing stdout :
Differing stdout :
where we are still evaluating why and where one of our recent commits
introduced a rendering bug (a small number of missing fills) in
example 16.
But the big news is we have just discovered how to implement
full-blown callback functions in modern Fortran with a generic
user-defined type included in the argument lists to carry all the
extra data that is sometimes needed in a callback. That model of
Fortran callback arguments (see the fifth kind of callback argument
processing in <http://www.fortran90.org/src/best-practices.html> is
exactly parallel to our current C practice and thus straightforward to
interoperate with what is done at the C level. So that has huge
implications for all Fortran API whose corresponding C API has
callback arguments (i.e., plcont, plshade, plshades, plmap*,
plmeridian, plslabelfunc, plimagefr, plvect, and plstransform).
Before, (i.e., plcont, plshade, plshades) where we simply had extra xg
and yg arrays as arguments and hooked that up with the C pltr1 or
pltr2 callbacks, we can now have a proper Fortran callback where the
user has complete control of what data is part of the defined type
argument. And before (the rest of the callback functions) where users
were very limited in what data could be carried in Fortran callback
argument lists, now they have the prospect of including any data that
they need in a user-defined type argument.
So due to this realization that full callback capability is available
for modern Fortran, we are going to end up with a much more powerful
Fortran binding of PLplot. And I don't think it will take me more than
a few extra days to get this all working.
Alan
__________________________
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); the Time
Ephemerides project (timeephem.sf.net); PLplot scientific plotting
software package (plplot.sf.net); the libLASi project
(unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
and the Linux Brochure Project (lbproject.sf.net).
__________________________
Linux-powered Science
__________________________
------------------------------------------------------------------------------
_______________________________________________
Plplot-devel mailing list
Plplot-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/plplot-devel
