Tools for unreserving a FW
remove use_global_spec option, and just have the parameters be taken from the root spec and overridden by the task parameters.
Speed up database queries
add max_runtime parameter to detect_lostruns()
Make it easy to set up the FW_config, my_launchpad.yaml, etc as environment variable. A separate tutorial, or maybe a single command? e.g. lpad setup
Store FW templates in “cloud”. Lpad has command to read/write template to Mongo, and templatewritertask can grab template from MongoDB.
FireTask name defaults to Class name.
Detect a blank config dir and give a proper error.
Allow FireWorks to block ports so that a parent job cannot override a setting. Maybe this is not needed?
Add stats, preferably using a MapReduce call for speed
Pitfall - putting the same FW in 2 workflows. Also note that RUNNING state updated a little bit after queue running state.
No negative fw_ids needed when returning FWAction
Go through logging, make sure it’s sane
Add option to automatically pass run dirs, e.g. send dict of {“fw_name”:”run_dir”} from parents to children
allow workflow connections to be defined by FW_name instead of by ids?
change docstring format to Google-style like pymatgen