One and only one HUSB?
Closed this issue · 2 comments
Page 38 says
_Note — The FAM record will be revised in a future version to more fully express the diversity of human family relationships.
Hopefully, this will change HUSB and WIFE to {0:M). Many same-sex couples today have both parties calling each other "husband" or "wife." Regardless of our opinions on morality, we need the ability to express reality without the mental gymnastics of “Sex, gender, titles, and roles of partners should not be inferred” or worse, “use several FAM records or use ASSOCIATION_STRUCTURE (p.45)s to indicate additional partners.” Polygamy was and still is practiced legally (or illegally) in parts of the world. If three men, three women, and several children live together and call themselves "family," it is neither realistic nor intuitive to require a family historian to document them in nine FAMs, especially when there is no evidence to determine which are the biological parents of each child.
For broader discussion, see https://www.webtrees.net/index.php/forum/8-translation/37317-spouse-versus-partner
Discussed in steering committee 28 Mar 2024
We agree that families in reality are much more complicated than 1 HUSB + 1 WIFE + n CHIL. We did not change this in 7.0 despite many rounds of discussion about other models because (a) many existing applications have hard-coded user interfaces with two parents in a given order (tree and fan views, notably) and the change was resisted by representatives of some of those applications and (b) we did not converge on a single alternative. We do expect it to change eventually, perhaps by removing the distinction between HUSB and WIFE entirely, though per semantic versioning this cannot happen any earlier than version 8.
We welcome suggested changes in the wording to make this more compatible with the diversity of family structures without changing the underlying data as a modification to 7.0.x or 7.1.
(a) many existing applications have hard-coded user interfaces with two parents in a given order (tree and fan views, notably) and the change was resisted by representatives of some of those applications
This is one of the reasons GEDCOM can’t get it right in many of the issue/problems it has. IMHO we are letting the inmates of bad design run the process toward good design!
A better design would not have any pointers saying HUSB or WIFE just a pointer to the individual_record and let the record tell us if they are male or female. In a relational database we would even have children too with a “junction record” to tell us how they fit in the family, as child, adopted, birth, parent, etc.
But bad design has prevented us from normalizing the data better, creating a true source_citation record, a place_record, shared facts, better name recording!