2.1.24: Merge pull request #314 from HSF/dev

optimize idds coordinator

2.1.23: Merge pull request #312 from HSF/dev

optimize iDDS coordinator to reduce the delay

2.1.22: Merge pull request #310 from HSF/dev

fix docker with postgresql

2.1.21: Merge pull request #308 from HSF/dev

(1) fix imports to be able to import from directory
(2) add functions to report async result status
(2) add functions to support pre_kwargs which can be used for global parameters
(4) optimize logging level (update several logging level from error to warning)

2.1.20: Merge pull request #302 from HSF/dev

update environment in the iworkflow
update stomp-py version requirement

2.1.19: Merge pull request #298 from HSF/dev

fix the asyncresult assumes return value 0 as not returning result

2.1.18: Merge pull request #296 from HSF/dev

(1)rename group_kwargs to multi_jobs_kwargs_list
(2) fix workflow wrapper, to fix the problem when single quotation mask is used to call the wrapper (in bash, when single quotation mask is used, bash will not interpret variables. It caused a problem to launch singularity container).

2.1.17: Merge pull request #294 from HSF/dev

(1) use pickle to serialize complicated objects such as tensor in ML (json cannot support it)
(2) to support presetup (split setup into presetup and mainsetup, to avoid the whole setup misleads some setting)
(3) fix stomp version (new version requires python3.11+)
(4) build container to support ax-platform, pytorch, botorch and so on

2.1.16: Merge pull request #292 from HSF/dev

fix return value from panda relay service

2.1.15: Merge pull request #290 from HSF/dev

Change ‘type’ to ‘workflow_type’. ‘type’ is a word kept in the python language. It raised some errors in the old python version.