[GRASS-dev] t.rast.aggregate test mysteriously worked one day

I’ve noticed that t.rast.aggregate “relative” test succeeded one day but generally fails:

http://fatra.cnr.ncsu.edu/grassgistests/reports_for_date-2015-04-27-13-32/report_for_nc_spm_08_grass7_nc/temporal/t.rast.aggregate/test_aggregation_relative/index.html
http://fatra.cnr.ncsu.edu/grassgistests/reports_for_date-2015-04-28-07-00/report_for_nc_spm_08_grass7_nc/temporal/t.rast.aggregate/test_aggregation_relative/index.html
http://fatra.cnr.ncsu.edu/grassgistests/reports_for_date-2015-04-29-07-00/report_for_nc_spm_08_grass7_nc/temporal/t.rast.aggregate/test_aggregation_relative/index.html

When I run it locally:

in grass session in a new mapset

cd temporal/t.rast.aggregate/testsuite
python test_aggregation_relative.py

I get these errors:

ERROR: Unable to make mapset element cell_misc/b_3
(/grassdata/nc_spm_testing/practice5/cell_misc/b_3): File exists

ERROR: Unable to make mapset element cell_misc/b_3
(/grassdata/nc_spm_testing/practice5/cell_misc/b_3): File exists

b_3 is the same name which is visible in (different) error messages in the tests:

ERROR: Unable to update raster map <b_3@__temporal_t_rast_aggregate_test_aggregation_relative>. The map does not exist.

ERROR: Unsupported relative time unit type for raster map <b_1@__temporal_t_rast_aggregate_test_aggregation_relative>: None

I looked to the test but I was not sucesful. If somebody would like to look at it, remember that tearDownClass class method runs just once after all test_ methods are executed while tearDown method runs after each test_ method. Also remember that the order of runs is undefined. Perhaps the error messages needs to be enhanced in the actual code.

Vaclav