@wxcafe True, False and Unknown

@cadey that doesn't answer why. why not just make another type that's 3 states and not call it boolean

@wxcafe ask the implementors of postgres?

@cadey @wxcafe you could do that with a nullable Boolean tho

or an integer with a range from 0-2

@Elizafox @wxcafe at the very least this makes things consistent across database applications

@cadey @Elizafox @wxcafe
No no, I can see the case for an unknown value. That's technically still consistent with boolian algibraic functions. I can't see calling an arbitrary thrinary list a boolian value though.

@cadey @wxcafe @Irick you misspelled Boolean twice

@Elizafox @cadey @wxcafe
I also mis-typed trinary with an h.

@cadey @wxcafe @Irick You did but that's dismissable as a typo

@Elizafox @cadey @wxcafe
It wasn't. it was intentional.

@cadey @wxcafe @Irick You sir have a poor sense of humour

I think you've hung around the functional programmers too long

@Elizafox @cadey @wxcafe
This is the sort of disparage up with which I will not put.

@Elizafox @cadey @wxcafe @Irick does that mean they spelled it right once?

@wxcafe @cadey OTOH is that still a Boolean? 🤔

@Elizafox @wxcafe A boolean by any other name would still find this statement to be false.

@Elizafox @cadey @wxcafe
Null isn't unknown though. We don't want to overload its meaning. We're not javascript.

@cadey @wxcafe @Irick Actually

in SQL

technically it is

@Elizafox @cadey @wxcafe

Null implies an absence of value. It's the Null set, yes? It's not, for instance, an uncollapsed qbit: that which has possible descrete value but has not been observed/recorded. It's a purposeful demarcation of 'absence' to a specific attribute.

But I suppose you could establish alternative interpretation with it.... AND IGNORE THE UNDERPINNINGS OF SET THEORY. D:<

@Irick @cadey @wxcafe Who gives a shit about set theory when you're not dealing with a set?

@Elizafox @cadey @wxcafe
...
What are relational databases but sets? Hell, even graph databases still rely on set operations.

@Irick @cadey @wxcafe Null applies to whole sets, not individual elements, iirc.

@wxcafe @cadey Many years ago I found that (this prior to the recent de-binarisation of gender), there were five values for "sex" in a major database system: male, female, unknown, indeterminate, and other.

Truth is complicated.