function.blah

Stream: general

[Reid Barton]

Did something happen where f.foo means function.foo f when f is a function?

[Reid Barton]

or was that just a suggestion?

[Gabriel Ebner]

No, I'm still waiting on a PR.

[Yury G. Kudryashov]

AFAIR, last thing happening was Gabriel pointing at the C++ source file that should be changed to achieve this.

[Reid Barton]

I was hoping we could rename the dreadful set.range to function.image--though I suppose we could do this anyways

[Patrick Massot]

Even better would be a trick to write f.map F where f is a function and F is a filter. Johannes sometimes wrote F.map f, I always try to write map f F

[Yury G. Kudryashov]

What should the parser to to understand that f.map F is actually filter.map?

[Patrick Massot]

That's tricky, it must use that F is a filter. Maybe :four_leaf_clover:

[Yury G. Kudryashov]

For filters you can probably use <$>

[Yury G. Kudryashov]

But this wouldn't work for bundled homomorphisms.

[Mario Carneiro]

there are universe issues with this alternative

[Patrick Massot]

I thought this never works because of universe issues

[Patrick Massot]

Oops, Mario was faster

[Mario Carneiro]

It works as long as you are on a known type, which is probably in Type, but anything generic wants to have two universes

[Yury G. Kudryashov]

I forgot about universes. Again.

[Patrick Massot]

Yes, that's one effect of the virus lock-down. It becomes even easier to forget about the outside universe.

[Johan Commelin]

Can the universe issues with <$> be fixed? Or are they inherent?