After a Smart Action in a Workspace is complete - The list (which is the source of that Smart Action) is updated with the requested fields = the ones show on the List.
Any field, of that collection which are placed in the Workspaces (outside of the list component) will not update at this points unless they are on the List component.
Note: If a smart field is requested on the List component - All fields are projected and so some Fields will update and appear to “normally refresh” but this is an artifact of the “remove projection” scenario of Smart Fields (returned for all flattened fields which are not smart fields)
Expected behavior
Workspace fields should update after Smart Actions are completed - or at least allow to specify those fields in the “refresh” response property of the Smart Action (only works for relationships today).
I have tried the same setup with forest-express-mongoose : “9.3.11” and it seems to be working on my side.
I tested running the smart action in a collection, and the update is reflected to 2 other component:
a label containing the value
the same collection in a different component
Could you please upgrade to the latest forest-express-mongoose version and test again.
If you still see the issue, can you please provide some more info about your workspace, settings, and if you see any frontend or agent errors.
Hi @Nicolas.M To make sure this is a Workspace thing and not related to the Liana version - I implemented a minimal repro on top of the new “agent” library (I hope this is good enough in this case, I cannot see why 9 vs 8 will fix this, but new agent vs 8 won’t).
From what I researched this is a design flaw in the Workspace feature and unrelated to server side API implementation - but I could be wrong of course.
Understood @nbouliol - No Problem.
While reproducing the above - at no stage did I have any console errors.
All HTTP requests (To the agent backend or Forest Api endpoints) completed with 200/204 - so no errors there as well.
Hopefully this moves us closer to a resolution.
Thanks for the added details, we managed to reproduce the issue.
We have opened an ticket on our side and started working on a fix.
We will keep this thread updated !