Closed hanfel-dovned closed 2 months ago
will studs ever work like marks? can you elaborate?
I think I'm being dumb with the mark comparison, but what I really want is something like
!<((stud-to-type stud.pail) vase.pail)
Actually, I'm realizing that Solution 1 won't work because of nested vases anyway. How do you guys feel about Solution 2?
I think I'm being dumb with the mark comparison, but what I really want is something like
!<((stud-to-type stud.pail) vase.pail)
This is intractable because of the hoon type system. There is a vague proposal to run certain files with a compiler that codegens $pail
to be a giant $%
of all the types it knows about, which would buy you what you want. But I guess I'm wondering what the use case is here?
Use-case: I ~zod
would like to have an item in my spreadsheet that is the longest message within my DMs from ~nec
I give my accel-cell a1 the dependency of /~zod/home/chat/~nec
The state of this cell is not very interesting to me, but its kids are interesting
What I want in the end is probably just a (list message)
/@ message /> con=message %+ turn ~(val by kids.ref) con
Superseded by #30
There are two solutions here:
Grab the entire
(map pith pail)
fromkids.bowl
, slop it in with the facekids
, and then have users manually get values from this map. Referencing kid values as a map here feels nicer to me thankidface.kids
, but it has the downside of the user not being able to de-vase the values using the stud of the pail, since they can't use ford runes inside of their code. The question here then is: will studs ever work like marks, with global availability? This would be a prerequisite to this solution feeling nice.Slop each kid's vase in individually and give them faces by parsing the
pith
into a@tas
. I think that this form of referencing is inferior UX due to its opinionated parsing decreasing the level of homointeractivity with the underlying system, but it is immediately doable, and @tiller-tolbus believes it to be superior UX (and I typically trust his intuition on this more than mine). Did we have another issue with this solution? If so, I forget what it was.Other ideas are welcome.