Added Parallel applicative and also breakingly...

[polytypic]

...changed template and apply to work in parallel and renamed
[upH|downH|h]asFailed to [upH|downH|h]asErrored.

[codecov-io]

Codecov Report

Merging #23 into master will increase coverage by 0.13%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master      #23      +/-   ##
==========================================
+ Coverage   99.15%   99.29%   +0.13%     
==========================================
  Files           1        1              
  Lines         237      282      +45     
  Branches       26       36      +10     
==========================================
+ Hits          235      280      +45     
  Misses          2        2

Impacted Files	Coverage Δ
dist/karet.xhr.es.js	99.29% <0%> (+0.13%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 74c0324...7fe41ac. Read the comment docs.

[abstracthat]

Hi @polytypic. I've finally reached the point of swapping out our app's API fetching layer and really need this Parallel algebra to pass to L.traverse, so that my XHR waterfall is not unecessarily serialized.

What do you feel is left before we merge this? I checked out the branch and the tests are passing. I've also added it to my work in progress branch in our app and it's just what I need. Happy to help get it shipped if you feel it needs anything else.

[polytypic]

Hmm... I don't recall anything particular. So, maybe review once more, merge, and publish if all looks good.

[abstracthat]

Sounds good. I've read through everything carefully and it looks good but I don't quite understand why we need the delayUnsub wrapper around the perform request. This causes the property to not be unsubscribed from until the next tick... do we need this for the merging the XHRs?

Any objection to releasing this as 1.0? It feels rather feature complete with these Happy path combinators. It's really pleasant to work with now that I've played with it and understand how to combine the requests with L.traverse. Swapping out the algebra is 💪

[polytypic]

I don't quite understand why we need the delayUnsub wrapper around the perform request. This causes the property to not be unsubscribed from until the next tick... do we need this for the merging the XHRs?

Yes, otherwise the subscription would be dropped momentarily resulting in repeating the subscription effect (abort and restart request). See here.

[abstracthat]

@polytypic I added apParallel to the readme since it's exported and could potentially be useful. If you are okay with the 1.0.0 version bump I'll ship this tomorrow.