Not sure if these are intentionally different, but I noticed that the number of children column in the people category view sometimes differs from the number of names listed in the Children gramplet. Specifically, for a person in my database who has no biological children and only one adopted child, the number of children is shown as 0, but the adopted child is listed in the Children gramplet. The families category doesn’t have a number of children column (why not?) but there also the adopted child appears in the Children gramplet.
If the difference is known or expected, that’s fine. Otherwise, if they should be consistent, I’m not sure which one is “right”. Just pointing it out in case others were unaware.
It would seem that Number of children is only counting children when the relationship to the parent is Birth, whereas the Children gramplet tab simply lists children linked to the parent’s family(s).
I am seeing the same thing for a cousin with 1 child by birth and 1 adopted child.
The Children Gramplet will also list multiple instances of the same person if they are listed under multiple households.
In the example.gramps database, Matthias Ball Jr had 3 marriages. In his first marriage with Abigail, I changed Jane to be his stepchild. (The Children column in the person view decremented from 9 to 8. And the Children Gramplet continues to list Jane.) Then copied Jane to the 2nd marriage (with Ann) and set BOTH maternal & paternal relationships set to stepchild. The Children Gramplet now lists Jane twice.
(This might be how I would log households if Abigail had been a young widow with a infant Jane. And where Jane died young and Matthias remarried. My reasoning to adding Jane to that 2nd marriage? If orphaned Jane was in the 2nd marriage household as a minor, listing her as a double stepchild would help avoid mistakenly creating a duplicate person for the next Census.)
The double listing would be reasonable if the Children Gramplet showed the Relationship to Active Person & to Spouse. It DOES show the clue of different spouses.
Further, the count in the Person view can be fooled into double-counting Jane if she is incorrectly set as birth child of both parents in both marriages marriages. This is a counting bug. The count should be of a unique person list.
I found and looked at the code. A child is added to the count if their relationship to both parents is “Birth”.
I tried a quick fix that added the child if the relationship to both parents is “Adopted” or to the father is adopted and mother is birth. Or the reverse, to father by birth and mother adopted. That took care of most of the children scenarios except a major segment of adoptions.
A great many adoptions are step parents adopting the children of their spouse by previous marriages. For the adopting parent, the count would be correct. For the natural birth parent, the adopted children get counted twice.
Absent a major rethink of the code, I do not see a quick fix. But for myself, looking at this issue, this was the first time I felt the need to enable the column.
So to fix the problem for myself, I just disabled the column.
For me, the number of children would be more useful in the Families category, and then there would be no issue of double-counting. Maybe two columns could be added to the Children gramplet: Relation to Father, and Relation to Mother? In that case, maybe it would be good if the count included all children in the family.