You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm curious why the design is to loop over the compiled releases in each transform? It seems less efficient than just doing that once.
Given that each transform calls run in __init__, it's not clear to me that a system of multiple classes is simpler than a single class (to maintain any global state) with one method for each transform (e.g. if the OCDS upgrade methods needed to maintain state between transforms, we could just put them all under one big class).
The text was updated successfully, but these errors were encountered:
Copying from f703eb9
I'm curious why the design is to loop over the compiled releases in each transform? It seems less efficient than just doing that once.
Given that each transform calls
run
in__init__
, it's not clear to me that a system of multiple classes is simpler than a single class (to maintain any global state) with one method for each transform (e.g. if the OCDS upgrade methods needed to maintain state between transforms, we could just put them all under one big class).The text was updated successfully, but these errors were encountered: