#rust-lang

aturon: I decided Conor McBride is the person most likely to know about partial functors: https://twitter.com/asajeffrey/status/882595822...

ajeffrey: associated types being able to do weird stuff is fairly well known

i.e. as in this case

in fact, when I opened your example, the first thing I started looking for was an associated type. Took me a while because I was scanning from bottom to top haha

nagisa: do you know of a reference for being able to code partial functors w/out HKT?

that is, it's totality that is the issue, not so much being a functor.

no references, no. I don’t tend to collect/remember references

nagisa: yes, AFAICT this is folklore :/

I'm hoping Conor knows a citation.

This is all a bit theoretical, mainly because it doesn't play well with type inference, e.g. a constraint Apply<OPTION, A> == Apply<OPTION, B> doesn't reduce to a constraint A == B.

AFAICT, associated types don't reduce quite when you expect them to, e.g. Apply<OPTION, A> doesn't always reduce to Option<A>.

when an RFC gets accepted, it often ends up getting harder to access it... the trouble is that people delete the branches they created for the PR; so the "Rendered" link in the first post of the PR becomes invalid

so, to actually read the RFC, I have to copy the RFC number, go to "Code", go to the rfcs folder, Ctrl-F, Ctrl-V, and then read that

yes

afaik it has become a semi habit to edit the rendered link in the pr description

to point to master

not sure where is the right place to document such trouble, but I feel like the situation could be better (e.g. by changing the "Rendered" link to link into the repo now)

right

but I dont have permission to do that; is there some place to put RFcs that didnt have that done yet?

or someone to ping to do that?

just ping someone with access

how would I know who does?

the best thing you could do is to collect a list of merged rfcs

that have the issue

and then post it in irc, someone will take care of it hopefully

well I have three that I was trying to read today:

you can also try pinging aturon or nmatsakis

rjung: Can't you still look at Files changed?

Yeah, so for the first one https://github.com/rust-lang/rfcs/pull/1857/fil...

Zoxc: yeah, and then the prettified view. that has annoying green bars though, and may even be outdated in the RFC had ammendments

Hello all, I'm not sure if this is perhaps too simple for this channel but I have a setup quite like this: https://is.gd/IAqU1O in my code. I was operating under the assumption that the trait bound `where` clause on TraitB would mean that wherever TraitB is used, that bound would be assumed. However, the compiler is making me repeat that same trait bound

everywhere TraitB is used.

Is my assumption incorrect or is this a bug?

Oh wow, it looks like after a full day of searching I find my answer immediately after asking. Looks like https://github.com/rust-lang/rust/issues/20671 might be the thing? Any progress been had on this or any workaround in the mean time?