2 ms·
The Lord Tensens gave me a new referent for what I normally think of as a mountain range: https://en.wikipedia.org/wiki/Tian_Shan https://en.wikipedia.org/wiki/
by 082349872349872 2y ago
The Lord Tensens gave me a new referent for what I normally think of as a mountain range: https://en.wikipedia.org/wiki/Tian_Shan https://en.wikipedia.org/wiki/Tian_Shan Vegetative cloning sounds almost like a shout-out to the The Island of Dr Moreau (1896)?
I'm aware of the other applications, but as I got to quantales by discovering what other people called the structure I'd gotten to via forcing distribution on informatics-friendly semantics*, rather than from another application, I'm still pretty ignorant of their wider use. In particular, my quantales are unital but not idempotent (but insofar as the pattern fragment also gives rise to a Heyting algebra, maybe I should learn more?) so that's why I haven't poked around with locales much. Maybe I should reread slides 18+ of https://people.maths.ox.ac.uk/~hap/tancl07/tancl07-abramsky.pdf https://people.maths.ox.ac.uk/~hap/tancl07/tancl07-abramsky....
* I want sequences (Hehner's "strings") to be noncommutative monoids, and alternatives (Hehner's "bunches") to be commutative monoids, and for them to distribute (in particular countably infinitely wrt to the Kleene star): AB* == A⊗∑B^i == ∑A⊗B^i (i running from 0,..ω)