Physics on the other hand very much has types, the unit to type correspondence is dopetastic.
Physics on the other hand very much has types, the unit to type correspondence is dopetastic.
Lately there's been a resurgence in foundations because of what new foundations like Homotopy Type Theory bring to the table. Funny how something as basic as equality can be revisited time and time again for thousands of years!
Was this decision ever explicitly expressed and agreed on by mathematicians at the time, or was it just implicitly adopted?
https://en.wikipedia.org/wiki/Curry%E2%80%93Howard_correspon...
But the connections don't really end there.
tl;dr: types are propositions and values are proofs of those propositions. Constructive logic is not concerned with (true xor false), but rather "I can prove A" or "I can prove not-A". There's also the middle ground of "I don't know how to prove anything about A".
But I preffer to keep that stuff in, even when doing pretty abstract stuff. Essentially because I wanted the type annotations. For example my favorite convention for the Minkowski metric is:
[c^2 0 0 0]
[ 0 -1 0 0]
[ 0 0 -1 0]
[ 0 0 0 -1]
That way your vectors are just $(t, x, y, z)$, where $t$ has units of time and the rest of space. Your covectors do come out oddly as $(c^2t, -x, -y, -z)$, but that's what makes them covectors.Another reason is that people instinctively feel that the elements of the coordinate tuple all have to have the same units. But that's not necessary when you have a metric around. Indeed keeping irregular units is helpful because it prevents you from accidentally forgetting to multiply through by the metric.