Following your suggestion and the instructions in Section 15.4.8.2 of the Octave manual, I changed the extension of the file opengl32 from bak to dll.  The problem has been solved.  Thank you.

I noted that my laptop uses Intel Iris Xe graphics, which is one I saw somewhere is known to have potential problems.

Thank you for your prompt response.

David

This is a question about the failure of Octave’s print command to create PDF files containing graphs produced with Octave 9.2.0 running on an HP laptop computer with the following particulars:

Processor:      12-th Gen Intel® Core™i7-1265U 1.80 GHz
System Type:    64-bit operating system, x64-based processor
Edition:        Windows 10 Enterprise LTSC
Version:        21H2
Installed on:   3/20/2024
OS build:       19044.4651
Experience:     Windows Feature Experience Pack
                1000.19060.1000.0

The installation of Octave is a default installation and was done in an administrator account, having selected an installation that  makes Octave accessible for all users.

Essentially, whether running in the GUI or the CLI,  Script A

# Script A
x=[0.0:0.1:10.0];
y=sin(x);
plot(x,y)
print -dpdf test.pdf

and Script B

# Script B
x=[0.0:0.1:10.0];
y=sin(x);
plot(x,y)
print -dpdf -debug test.pdf

both create the plot but both fail to create the requested PDF file containing the graph.

The GUI responds differently than the CLI, but both hang on the print statement and neither produces a file containing the figure. Run in the GUI and the CLI, these scripts produce different results/information, but I am unable to interpret that information. I hope the information below and the attached screen copies will be enough for you to figure out why Octave’s print statement fails to create the requested file.

Script A run in the GUI

Execution of the print statement causes the system to freeze and be unresponsive both to <CTRL-C> and to clicking the left mouse button on the ‘X’ in the upper right corner of the window.  No error messages appear in the GUI, and the requested file is not created. The execution can be halted only by stopping the task with the task manager or by waiting until the OS asks if I want to close the program or wait until it responds.

Script B run in the GUI

Octave displays the message in the attached screen copy GUIdebug.jpg (though the text runs off the screen and I can’t figure out how to capture that part).  Otherwise, the behavior is the same as that described for Script A in the GUI.

Script A run in the CLI

The script displays the message

        GL2PS warning: Unknown token in buffer

and the program hangs. Typing <CTRL-C> generates the error message in the attached screen copy CLInodebug.jpg .  In this error message, it appears as if the system fails to create the file oct-pdnWVD in my AppData\Local\Temp directory.  Evidently, the program ultimately needs this file but can’t find it.  (I checked in the directory and the file, in fact, is not there.)

Despite the <CTRL-C>, the program is hung and does not return to the Octave prompt.  I have to close the window by clicking on the ‘X’ in the upper right corner to escape from Octave.

This time, the file test.pdf (attached) is created, but it contains only what I assume are header and footer for the missing description of the graph.

Script B run in the CLI

The script provides the message in the attached screen copy CLIdebug.jpg before hanging and creates the file test.pdf (not attached; it appears to be the same as the PDF file created by Script A in the CLI.).  This time, the file of interest in my AppData\Local\Temp folder is oct-d7xtCg.ps, which file is created.  Its entire contents is

%!PS-Adobe-3.0
<< /Margins [0 0] /.HWMargins [0 0 0 0] /PageOffset [181 302] >> setpagedevice
%%EOF

Similar files appear to have been created in the other runs in this message, but they were not mentioned in the error messages.

I would be grateful for any wisdom you can provide to fix my problem with Octave’s print command.