plot package: 1 possible bug and 1 question

[cae]

Hi.

Apologies if these 2 items have been discussed already or are wrong.

1) I find the gfortran fails to compile the slevels.F routine in the plot/src/Utility directory, with the variable tindx already defined.

Compile works if I change line 35 as follows:

Code: Select all

      
      ! integer file, nc_gid, tindx
      integer file, nc_gid

2) I find that the year set by the 1st line of the input file (e.g., for ccnt) is being ignored. For example, with the file WC13/plotting/ccnt_i4dvar_increments.in (which is part of the test directory on the svn server), the output gmeta file does not change if I modify the first line from

Code: Select all

2004 0   :  year and starting year-day (use yearday<0, for no time label)

to

Code: Select all

2002 0   :  year and starting year-day (use yearday<0, for no time label)

Judging from the comments in the input file and the plot wiki, these two numbers should be the starting year and year-day for the plots, so I would expect to see different dates in the resulting plots.

My recollection is that this number had the expected impact in previous versions of ccnt, but I'm having trouble reproducing them because of changes to the package.

Note that the second number, the year-day appears also to be ignored if it's >=0.

As a related suggestion, could ccnt automatically determine the proper date as ocean_time has the relevant initialization time as an attribute?

Thanks very much.