Reported by @AnsleyManke on 9 May 2003 17:56 UTC
Prior to v5.52, the sequence SET GRID/SAVE; CANCEL DATA/ALL;
SET GRID RESTORE gives a final result of the last SHO GRID
as %%. It should be the grid of ROSE. Example with notes:
yes? use etopo20
yes? set grid rose !<== this should bump the use count on this grid
yes? sh grid
yes? set grid/save ! <== this should bump it a second time
! (Note that the previously stored grid
! may need to have grid_use_count decremented)
yes? use geo_borders
yes? set grid CONTINENT_LON ! <== this should decrement the
! use count on the "rose" grid
yes? can data/all ! <== the use counts on the "rose" grid is
! still non-zero, so it will not be deleted
yes? set grid/restore ! <== this should increment the use count
! as a result of SET GRID
yes? sho grid
Check whether the SET GRID/RESTORE should also decrement the
same use count because the save/restore grids are on a stack,
and this command pops that stack. The net result of these two
actions is that the use count remains unchanged. (May require
just a comment in the code)
Reported by @AnsleyManke on 9 May 2003 17:56 UTC Prior to v5.52, the sequence SET GRID/SAVE; CANCEL DATA/ALL; SET GRID RESTORE gives a final result of the last SHO GRID as %%. It should be the grid of ROSE. Example with notes:
Check whether the SET GRID/RESTORE should also decrement the same use count because the save/restore grids are on a stack, and this command pops that stack. The net result of these two actions is that the use count remains unchanged. (May require just a comment in the code)
Migrated-From: http://dunkel.pmel.noaa.gov/trac/ferret/ticket/546