Interesting Results (possibly)

#1 · 2025-01-17, 6:54 am

I noticed in my main database the person table has people with flags (>0) but where both Relate1 & Relate2 = 0.

Most of them do not show and obvious why other than they may have had a connection (or should have had ) to a common ancestor (known or unknown) but wonder how or why RM is giving them flags

Script #1

shows counts where Relate1 + Relate2 = 0 but flags <> 0 by totals of each flag

SELECT flags, count(PersonID) as qty
FROM PersonTable
WHERE Relate1 + Relate2 = 0
Group By flags

Script #2

where Relate1 + Relate2 = 0 but flags <> 0 by person

SELECT PersonID as PID
FROM PersonTable
WHERE Relate1 + Relate2 = 0 and Flags > 0

#2 · 2025-01-17, 8:28 am

Interesting but, hopefully, of no consequence. I surmise that these people are truly not in the same tree as the focus person for the latest Set Relationships but were for an earlier but different focus person. RM then does not wipe the entire Flags column on a new Set Relationships but updates on those who are spouses of directly linked people.

One could test the Kinship List for possible consequences and assume its results would correspond to those of Set Relationships by doing a document comparison in Notepad++, Word... of the reports before and after a SQLite reset of Flags.

kevync has reacted to this post.

#3 · 2025-01-17, 8:55 am

Quote from Tom Holden on 2025-01-17, 8:28 am
Interesting but, hopefully, of no consequence. I surmise that these people are truly not in the same tree as the focus person for the latest Set Relationships but were for an earlier but different focus person.

actually is a good hypothesis. have to do some more testing...to figure out connection.

#4 · 2025-01-17, 10:28 am

My Approach was probably simpler --

UPDATE PersonTable
Set Flags = 0
WHERE Relate1 + Relate2 = 0 and Flags > 0

then go back in RM and set relationships.

The people who were no longer related did not get flags back -- I would call that a design flaw that the flags are not reset when relationships set. I guess I could do some more testing to replicate the whole but maybe someone has some better examples. I am going to report "bug" on RM community board.

#5 · 2025-01-17, 11:44 am

Is it a "bug" if there is no adverse consequence to users of the RM application? Have you identified any?

It may be a factor that SQLite exploiters need to take into account for certain queries but that wouldn't be much of a motivating factor for RM to fix, given the many things from us that have been ignored.