Intel® Fortran Compiler
Build applications that can scale for the future with optimized code designed for Intel® Xeon® and compatible processors.

PGPLOT problem with IVF10

john_humble
Beginner
703 Views

For a long time I've been using a plotting routine based on PGPLOT.My routine iscalled by several programs (using QuickWin)and can output to either a screeen or, via postscript, to a printer. All this workedwith CVF and Win 98, ME & XP, with one peculiarity. When asked to produce postscript output the programs would briefly flash a blank windoweach timethe plotting routine was opened. The window immediately closedwithout user intervention so I didn't bother to try to track down the cause.

Recompiling and rebuilding everything (includingPGPLOT.LIB) using IVF10 and WinVistahas brought a problem. When postscript output is specified for one of my programs the blankwindow now refuses toself cancel. Instead, it stays on the screen until the user closes it via the X button. A useless error report appears (Error 0) and the run continues.

I've dug my way back through the PGPLOT Fortran, so far without success. I've also emailed Tim Pearson, but he's not replied. This time of year he may be on vacation, or he may just have abandoned maintenance. Does anyone on this forum have anyexpertise in this area, or can suggest someone or some group which may be able to help?

Alternatively,does anyone have a recommendation for a package that might be able to replace PGPLOT without too much effort?

TIA John

0 Kudos
4 Replies
Steven_L_Intel1
Employee
703 Views
The blank window is probably caused by a SYSTEM library call, though I have not heard of this behaving differently on Vista. I'd be surprised if this was specifically an IVF issue - what happens if you run the same EXE on XP?

Are the sources for PGPLOT provided? One might try replacing the SYSTEM call with a use of the Win32 API routine ShellExecute.
0 Kudos
john_humble
Beginner
703 Views
Thanks for the suggestion. I have the PGPLOT source code. Had to modify it slightly to compile with IVF, which wouldn't allow a couple of usages of dummy arguments within subroutines which CVF hadn't complained about. I doubt very much whether that's the problem. I'll experiment with reverting to the CVF compiledPGPLOT.LIB, and alsowithreplacing the SYSTEM call. I'll not get back to the job until later in the week
0 Kudos
Daft_Pict
Beginner
703 Views
Thanks for the suggestion. I have the PGPLOT source code. Had to modify it slightly to compile with IVF, which wouldn't allow a couple of usages of dummy arguments within subroutines which CVF hadn't complained about. I doubt very much whether that's the problem. I'll experiment with reverting to the CVF compiledPGPLOT.LIB, and alsowithreplacing the SYSTEM call. I'll not get back to the job until later in the week

John, I found this thread whilsttrying to resurrect some old vms fortran code and it happens to call PGPLOT. I downloaded the last version of the source I could find (v5.2.2) and tried to build it as a DLL to link my code against. I haven't been successful yet and wondered if you could share the changes you made. With 2 of us trying to make this work we may have more luck! TIA &Cheers - Brian

0 Kudos
jimdempseyatthecove
Honored Contributor III
703 Views

For a long time I've been using a plotting routine based on PGPLOT.My routine iscalled by several programs (using QuickWin)and can output to either a screeen or, via postscript, to a printer. All this workedwith CVF and Win 98, ME & XP, with one peculiarity. When asked to produce postscript output the programs would briefly flash a blank windoweach timethe plotting routine was opened. The window immediately closedwithout user intervention so I didn't bother to try to track down the cause.

Recompiling and rebuilding everything (includingPGPLOT.LIB) using IVF10 and WinVistahas brought a problem. When postscript output is specified for one of my programs the blankwindow now refuses toself cancel. Instead, it stays on the screen until the user closes it via the X button. A useless error report appears (Error 0) and the run continues.

I've dug my way back through the PGPLOT Fortran, so far without success. I've also emailed Tim Pearson, but he's not replied. This time of year he may be on vacation, or he may just have abandoned maintenance. Does anyone on this forum have anyexpertise in this area, or can suggest someone or some group which may be able to help?

Alternatively,does anyone have a recommendation for a package that might be able to replace PGPLOT without too much effort?

TIA John

Create ashortcuttothe program that is producing the blank window (using Windows Explorer right click on program and select Create Shortcut) then rename the shortcut to suit your needs. Then right click on the shortcut and select properties. In the properties you can select to run in minimized window (Shortcut property page, Run pull down selector). If that does not produce suitable results then pre-pend in the program name in SYSTEM("program args") a call to the command shell with /C

SYSTEM("CMD /C program args")

or try

SYSTEM("START /MIN program args")

Jim Dempsey

0 Kudos
Reply