Ditch httr in favour of crul

[maelle]

I'd like to ditch httr in favour of crul. This way we could use webmockr as well (part of the @sckott -verse)

It wouldn't be too much work since for the simple requests here using crul is no rocket science. I use it in http://github.com/ropensci/ropenaq now and at the time the difficulty was understanding asynchronous requests which we wouldn't need here.

@dpprdan @jesssadler would you be up for learning crul? Otherwise I can do the translation, this doesn't change the design which you both work on.

After opencage next release then an rOpenSci blog post (or 2) about making a package user-friendlier + testing it better would be interesting I think 🤔

[dpprdan]

Apart from your enthusiastic blob post, I don't really know crul? What does it do better than httr (apart from async requests, which is not relevant here)? I'm open to it, but don't see the benefit, yet.

[maelle]

It works with webmockr ☺

[sckott]

it will also work with vcr very soon to cache http requests for test suites

[dpprdan]

Isn't that what httptest does (as well)? But I don't really know what I am talking about, so, yeah, go ahead. 😄

[maelle]

I do have a bias for rOpenSci tools 😂

[jessesadler]

I have no real opinion or any expertise in this issue. Trying to learn as much as I can about this stuff, but I am not too far yet. ☺️