https://ghostscript.com/pipermail/gs-devel/2012-February/009149.html

>> In order to flush the device commands to the "page", use the .outputpage
>> Ghostscript
>> operator.
>>
>> For example: 1 false .outputpage
>>
>> The '1' is a copy count (not really used) and the 'false' tells
>> Ghostscript to not erase the
>> buffer, similar to the Level 1 and 2 'copypage' action.
>>
>> I hope this helps
>>
>> Best Regards,
>> Ray Johnston

.outputpage is an internal command whose use is not recommended.

and apparently the above works differently in different contexts.

-----------------------------------------------------------------------------
Reading, e.g.
    /usr/share/ghostscript/9.22/Resource/Init/gs_init.ps
is generally very informative 

# testing sample:

srcdir=/home/gbc/HOPS/trunk/sub/dfio/copypage \
GS_DEBUG=gs_debug.ps \
GS_COPYPAGE='PGPLOT restore (1\n) print flush flushpage (2\n) print flush prompt showpage' \
FP_SLEEP=-20 ./fplot_test


-----------------------------------------------------------------------------
https://www.adobe.com/content/dam/acom/en/devnet/actionscript/articles/PLRM.pdf

-----------------------------------------------------------------------------
Email to tmorin Aug 8 2018,

My machine at home is also gs --version 922.  Comparing the guts
of the Resource/Init directory, there were a couple of minor files
(fonts) different and 

/usr/share/ghostscript/9.22/Resource/Init/gs_init.ps

parsec:

 87422 May 30 17:27 /usr/share/ghostscript/9.22/Resource/Init/gs_init.ps
  87977 Apr 23 09:01 /usr/share/ghostscript/9.22/Resource/Init/gs_init.ps

  The difference is /UndefinePDFOperators which list operators not needed
  after startup.  Parsec listed /flushpage, my machine didn't.  Apparently
  the concern here is people hacking unused operators....

  I removed /flushpage from the list and parsec is now fine.

  I can't wait to see what happens in 9.23.

-----------------------------------------------------------------------------
# eof