I have a modestly complex data model, from which I'm using a table visual to flatten out various categorical data fields. The end result should be a row with an ID number on the first column and all kinds of various text fields any of which (including the ID) could possibly be empty.
This should be fairly straightforward; however, I'm finding columns to the right of a blank don't fill in, so if the blank is the ID number on the far left, the whole row disappears. If I move the ID number to the far right, it shows blanks as expected. I did the same with a variable from another table that also would mismatch and yield blanks, it exhibits the same behavior.
So, what I expect:
001 A B C
V E X
002 B O T
What I get:
001 A B C
002 B O T
Or:
A B C 001
V E X
B O T 002
I've tried setting all variables to "show items with no data," no luck. I confirmed there are no unexpected filters affecting the table. I tried using a measure to fill in an "x" when the ID number is blank but I can't get that to work because the blanks are artifacts of the mismatched data, not a null cell. All my connections are 1:M and bidirectional.
The only clue I've found is that it was built in an older "for report server" version, and when I built the same schema with arbitrary data in the current Desktop version, I can't recreate the issue. So what I'm hoping is it was a bug that was fixed in the last year or so. Otherwise it might be a setting somewhere I don't know about, or some quirk of the real data. Can anyone narrow down those possibilities?
