@samth Did you have in mind more for deffl? The flids identifier is not used

Also, I think it would be better to use a long and descriptive name in place of deffl, since it’s only used once and there’s just once namespace for all “core-*.ss”

Nit: I much prefer to put a newline after begin (when the whole form isn’t on one line) and cond

@mflatt Oh, I should have removed that once I realized that flsin et al were already defined but not provided

I’ll make those changes

Thanks!

What do you think the right thing to do with the various unsafe-fl... functions is?

The ones that are not currently provided by #%unsafe? Eventually, it will make sense to add them to #%unsafe so the Chez unsafe versions are made available

Ok

With that PR, racket/promise now demonstrates what looks like an expander-level bug

Yes, I have been looking at the same thing :disappointed:

I think it’s related to the struct layer, but it will be a while before I can investigate more

Well, actually, I was looking at a different problem related to “expander/expand/root-expand-context.rkt”

Is that the apply-syntax-parameter error?

I don’t think so, but I’ve pushed repairs that let racket/promise load (including the same changes that you made, I think)