Hi, I noticed that baserow displays more than one category for “empty” in column/field summary for distribution.
what do these actually correspond to?
this might be a bit of a problem for users as there is no apparent way of filtering for these.
seem to be related to the distinction: null vs empty string.
in N8N this shows up as either null or “”.
This is strange indeed. I tried to reproduce it on my end and couldn’t. I’ll do further tests and create a GitLab issue for this. I suspect it is the difference between a NULL and an emptry string as you suggested was the case in n8n. Just need to work out why or where the difference is being created.
Thanks @dev-rd
I can’t reproduce this on my end. We suspect this might be from a bug that was already fixed (but we couldn’t fix existing fields due to risking data loss).
@dev-rd Could you try duplicating the field and seeing if this distribution issue still exists in the duplicated field?
The issue still exists for duplicated fields (Baserow 1.33.4)
I was able to reproduce the issue for a new table in Baserow 1.33.4 for duplicated fields converted from single select fields back to text
the key step would be to delete an item after converting from single select back to text.
If you delete any field after such conversion it creates the second empty category (by delete I mean hit the delete key)
I’ve worked out the issue and created it on gitlab here: When adding a new single line text field with existing records will result in NULL values instead of empty strings "" (#3668) · Issues · Baserow / baserow · GitLab
thanks - btw this also seems to carry over to dashboards as two separate items in graphs.