Rose suite validation fail reporting

[hjoliver]

A NIWA user has suggested (and I tend to agree) it's confusing that on validation fail rose suite-run prefixes all cylc validate output with [FAIL], even the lines that are merely warnings or info:

[FAIL] cylc validate -o /tmp/tmpZGmm8o --strict nzcsm_L70_40km # return-code=1, stderr=
[FAIL] 2020-05-20T01:10:23Z WARNING - Queue configuration warnings:
[FAIL]  + ops_queue: ignoring 'OPS_1PE_1' (task not defined)
[FAIL] 2020-05-20T01:10:23Z WARNING - Queue configuration warnings:
[FAIL]  + ops_queue: ignoring 'OPS_1PE_1' (task not defined)
[FAIL] 2020-05-20T01:10:23Z WARNING - Queue configuration warnings:
[FAIL]  + ops_queue: ignoring 'OPS_1PE_1' (task not defined)
[FAIL] 2020-05-20T01:10:23Z WARNING - Queue configuration warnings:
[FAIL]  + ops_queue: ignoring 'OPS_1PE_1' (task not defined)
[FAIL] 'ERROR: No cycling sequences defined for forecast_da006_done_1'

(Only the last line is an actual "fail").

[hjoliver]

I know this output is not technically incorrect, and users who look closely and think about it a bit may understand exactly what it means ... but still, we could probably make it more obvious?

[oliver-sanders]

The FAIL is due to the cylc validate return code, would have to grep log messages to do anything more advanced, or better still get rid of rose suite-run ala Cylc8.

Suggest we close this with Rose2/Cylc8.