It isn’t. The Social Security Administration tracks births and deaths, but they don’t track marriage. When your spouse dies, they don’t automatically send you a check; you have to ask, and provide proof of legal marriage. The real database is millions of paper marriage certificates in millions of homes.
I worked at the IRS for a while, and was surprised to notice there’s no check-box for gender on the tax form. Doesn’t matter if you’re he, she, or something else. Although the IRS didn’t then recognize same-gender marriage, if two men filed MFJ it would probably zip right through the computers and here comes your refund check. The only time it would cause problems is a full audit, which is really very rare. This is not because of the repeal of DOMA, it’s always been this way.
So I say the Y2Gay problem is imaginary, except in the form of the prejudices of a thousand county clerks, which is very very real.
The choice of quoted text is odd, since it’s not about gay marriage, its a hypothetical exercise in thinking about how to handle plural marriage, where (for example), Alice and Bob marry, then Carol joins in, and later Alice leaves.
It looks like the poster just pulled out the most confusing bit of text, without understanding what it was about.
This becomes significantly more complicated in dealing with children. For instance in a school situation you could have Jimmy’s biological dad Ron, his Biological mom Carol, his grandfather paternal Gary, his Step mom (Rons second wife) Jane, Janes second husband Carl, his Psychiatrist, his social worker, his adult brother who has custody, Carol’s wife, and so on. Then you may have notes that Ron cannot pickup Jimmy because of a restraining order and Janes second husband is the preferred contact who basically raised Jimmy, but is now married to someone else who would be what-- his step moms ex husbands new wife. And the system needs to red flag anyone that shouldn’t be allowed contact, and this can have effective dates. Or they could all have a mix of custody and the teacher better not call the wrong person.
It’s interesting to note that the IRS has never had this problem. Take a look at the 1040. All meatsacks are nothing more to them than a social security number - filing single, dependent, married (joint) or married (separately). They don’t give a f@#$ what your gender is as long as you cough up your goddamn money or we’re going to have to hurt you.
From the computer and database schema perspectives, there was never any need for this kind of gender limitation. It’s just extra (and wrong) pointless work. Some hicks coded in unnecessary checks like all years have 365 days (wrong), everyone has a first middle and last name (wrong), you can divide everyone into male and female (wrong), everyone has two parents (wrong), etc.
Assumption of monogamy? I was thinking more of controlling the situation where someone signs up for spousal benefits, toward which the employer makes a contribution, for forty-'leven people in a plural marriage. The idea behind the restriction was purely economic.
And this is why my plan to just assign all the users either GUIDs or MAC addresses will eventually succeed.
Now they say ‘too hard to remember’, ‘too unfamiliar’; but after a few generations of culturally unhindered drift, the state of anthroponymy will be so convoluted that they will beg, beg!, for the simplicity of my plan!
The situation is, indeed, kind of irksome; but student information system software, because of the need for supporting “Who is allowed to pick kiddo up?” “Anyone we should call the cops on if they try to pick up kiddo?” and assorted similar cases has largely adapted through generalization:
I can’t speak for all systems, of course; but the ones I’ve seen basically adopt the “We don’t care about your network of social relations, we just care about how you are associated to the student”. This allows you to ignore most of the complicated backstory and just divide the world into:
Custodial parent(s) and/or guardians: Can pick kiddo up, are default attempted contacts in the event of illness, accident, parent/teacher conference, etc. Receive report cards and generic school mailing list chatter, are typically entitled to access to FERPA covered records unless student is 18 or older.
Noncustodial parents and/or guardians: Generally not cleared for interaction with kiddo at school unless specifically noted; but may be authorized to receive report cards and FERPA covered records.
That’s pretty much it. Doctors, psychologists, and similar don’t typically go through the school for access(just as prospective employers don’t go to the school for transcripts, the student or parent/guardian requests those records and passes them on if they wish); social workers or other state agents might gain direct access through a legal process of some flavor.
You can’t beat the fact that any database design that copes with the real world will simultaneously be overkill for 95% of cases and supplemented by human-readable-only special instructions for issues not amenable to the database in the other 5%; but KISS is a very, very, valuable heuristic.
Never start from examining the ‘blooming, buzzing, confusion’ and freaking out. This approach is accurate; but hopeless. Instead, start from the limited set of operations that you actually perform(‘release student to person picking them up’, ‘mail report card’, ‘authorize access to FERPA-covered records’) and put the world into piles based on those. Reductive; but relatively robust.
It’s also assuming that all members of said plural marriage become joined to all other members of the marriage. While that may happen, it’s the minority of the cases of polyamorous relationships. I’m having bad flashbacks to one DBM class which used Prolog. Trying to define familial relationships actually hit more than a few cultural snags. (Instance: the professor absolutely insisted that the sibling relationship required both parents to be the same. I finally gave up and just defined a half-sibling relationship, but left my snarky comments in the code.)
Well, it’s mostly just a thought exercise, I think. Illustrated by the following:
That, and the fact that it starts wandering off into polyamourous relationships at the end. I don’t think it’s seriously positing that there’s going to be a huge Y2Gay database crisis.
