Thank you for your analysis.
Indeed that brings back consistency. So, I guess the best practical solution to this problem is to always include the above statement in your plotting script.
yes? go try2
yes? show symbol *len*
PPL$XLEN = "5.60900"
PPL$YLEN = "8.46800"
yes? go try2
yes? show symbol *len*
PPL$XLEN = "4.74200"
PPL$YLEN = "7.15900"
What are these lengths? When I compare the two plots side by side, like so
yes? go try2
yes? show symbol *len*
yes? set window 2
yes? go try2
yes? show symbol *len*
the value of PPL$YLEN is different between the two. In addition to the "JAN" vs "J" difference, the sizes of other letterings are also different. But the sizes of the axes and colorbar seem to remain the same on my screen!
The latter sizes are specified by my viewport relative to the (logical) size of the window . . . That means that the size of the window is changing between the two calls to try2.jnl . That means that the "set viewport" command somehow changes the size of the window . . .
Regards,
Ryo