I've been looking at easy to fix bugs on the bug tracker and noticed the
one regarding plsmema in the f95 bindings.
plsmema (and plsmem) are marked as part of the Common API, but they are
not consistently propagated to all the bindings and there is no
standard example to test them. The only example using them is a specific
I've added support for C++, f77 and f95, though I'm not sure how the
fortran support will work, and certainly the f95 interface needs fixing
up (there was already a comment about this for plsmem).
It would be nice to have a standard example to demonstrate the use of
these functions and to test they work. I would propose that a simple
way would be to use one of the existing examples then use plmkstrm,
plsdev, plcpstrm, plreplot to save the plot to memory. Could do what
the python script does and save it as a png file (or other simple
raster format). This would mean that it would still work as a
"standard" example, but would produce some additional output. It would
look very similar to what happens with example 1 at the moment.
Longer term, plsmem supports the mem, memqt and memcairo devices, but
plsmema support is only available to the memqt and memcairo devices.
Adding support to the mem driver would be nice, and not necessarily too