A few weeks ago, at the time of the most recent PLplot release, I posted a short announcement on comp.lang.ada, and I gave a very simple example of usage in an Ada program. Someone asked why plinit couldn't be done in the (Ada) elaboration of PLplot and why couldn't plend be done in the (Ada) finalization when the program terminates. These are Ada concepts meaning roughly that plinit could be worked into Ada's own initialization (aka elaboration, sort of) and finalization and thus executed automatically. I replied that in some usage situations PLplot needs to do some work before plinit is called. So that left the question of whether it made sense to put plend into an Ada finalization stage. I remarked that the manual method currently used would be better if the user wanted to free resources before the program was done running, but that it might be a good idea to have the automatic method (in the Ada finalization stage) in case the user forgot to do it manually. Then someone pointed o ut that if the program raises an exception, the manual method would be bypassed. We could ask the user to write an exception handler but s/he might forget to do that too.
One might consider removing the manual method (and keeping backward compatibility by making the Ada-level call do nothing) but I don't like removing the manual option for reasons stated above. So before trying to put in an automatic call to plend at the Ada finalization, I need to know: (1) Does it hurt to call plend twice without an intervening plinit? I.e., the user puts in a manual plend and Ada does another automatic plend. (2) If so, is there a way to find out if PLplot is "active," that is, has there been a call to plinit but no matching call to plend? Jerry ------------------------------------------------------------------------------ October Webinars: Code for Performance Free Intel webinars can help you accelerate application performance. Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from the latest Intel processors and coprocessors. See abstracts and register > http://pubads.g.doubleclick.net/gampad/clk?id=60135991&iu=/4140/ostg.clktrk _______________________________________________ Plplot-devel mailing list Plplot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/plplot-devel
