Change: address #34 by updating combine*'s API

[rpearce]

Closes #34 by addressing the issue there for all combine* functions.

The tl;dr is that the data needs to be sent, by itself, as the last function call, similarly to how converge and juxt work in ramda.

You can call combineWith and combineWithP like (f)(g)(a) or (f, g)(a), but not (f, g, a) nor (f)(g, a).

IMO, this constitutes a breaking change and is a major version bump unless y'all can somehow justify it as a bug fix.

Let me know what y'all think and what changes you want to see!

[mgreystone]

I think useWithP also has this problem. I'd doublecheck all the assemble*'s too, though a quick glance makes me think that it might be fine.

These look perfect to me.

[rpearce]

useWithP and assemble* all look 👍 to me

[mgreystone]

We'll need to make it clear this is a breaking change (on purpose this time).

[rpearce]

I don't see a CHANGELOG.md for us, and there is only 1 release.

I can do a GH release and write up the info there for this latest version once this is merged.

[rpearce]

I'd like to wait for another ✅ before merging. @mgreystone is there anyone else that you think we should request review from who would be interested?

[mgreystone]

I believe rise.com is the largest consumer of funky. cc @jasminabasurita @JWA111

[rpearce]

Going to merge this on Monday and publish a major version then.

[rpearce]

No additional feedback. Merging.