a bit surprising:

(s/valid? (s/map-of keyword? int? :kind vector?) [[:a 1]])
=> true

what is that even supposed to mean?

seems like it's doing exactly what the docstring says to me

I don't know what that's supposed to mean, I'm trying to build a tool on top of spec and I have to support what it supports

I'll probably ignore this particular thing... I just was trying to understand what kind of data shapes various variations of s/every support.

spec needs a spec? ;)

less surprising:

(s/valid? (s/every-kv keyword? int?) #{[:a 1]})
=> true

what kind of tool are you building, or is it secret?

"spec forms"

something cljfx related?

It's a reveal plugin that builds a UI form for a given spec. Then you can fill the form to get back the data in the shape that satisfies the spec.

https://twitter.com/v1aaad/status/1390689740308074498?s=19

...and the form resembles the data

so it's way to visualize a spec form?

More than that. You can bind changes in form to custom code and custom views. For example, if you have spec for vega/vega-lite data format, you can have a UI form for creating vega chart input and a vega chart itself that updates on changes to the form. This allows learning and experimenting with data-driven APIs interactively

every-kv is really about "seqs to map entries", but we have no way to say that in Clojure (yet)

Rich and I have been talking about that gap for years, some day we'll figure it out :)

what's the 2-version of triple?

duple?

map-entry?

pair is commonly used though :)

duple-blob

duple-flock

solved? ;)

the "seqs to" part is the hard part :)

btw I think every-kv is fine, I'd expect it to accept any coll

but maybe the docstring shouldn't mention associative...

I think that may be vestigial from earlier impls