Ronnie Dutta
Ronnie Dutta
I tried your example running on 8.0.x and this branch, but I can't say I noticed a difference in the changes to the `task_states` table
Oops, this breaks the ability of Cylc to restart `8.0.{0,1,2}` workflows. ``` ERROR - no such column: task_states.is_manual_submit Traceback (most recent call last): File "~/github/cylc-flow/cylc/flow/scheduler.py", line 670, in start await...
(Was trying to get the tests to run)
(To be merged after #4274)
> Wait on fixing the DB json key hack from SoD Was this #3863?
A new logging level between `INFO` and `DEBUG` would be useful for `cylc clean -v`. That would tell you what exactly files are being deleted without any of the other...
If we shorten `DEBUG` I would prefer `DBUG`
The difference between `initial cycle point` and the other cycle point settings in `*-start.cylc` would appear to be due to this line: https://github.com/cylc/cylc-flow/blob/09581beb94250374215bea925f2bec985f6c13c2/cylc/flow/config.py#L273-L276 This is done before the dense config...
> Related to this is the fact that `cylc config` accepts `--icp` as an option but none of the other cycle point settings which seems rather strange. That seems like...
Is this about trying to do something like this? ```shell create_test_global_config "" " #!Jinja2 [scheduler] process pool timeout = {{ X }} " ``` If so I think the simple...