when I tried to demonstrate the problem to someone else here at Fermilab,
I wasn’t able to do that - I placed a few buttons on a frame and successfully saved the result into a ROOT macro…
Not sure what resolved the issue - apparently there was something of an operational nature I didn’t control. An anomalously huge NFS delay could give an example of an explanation - I was writing
to an NFS-mounted disk .
This topic can be closed. Thanks again for spending time on what happened to be a red herring.