Fork me on GitHub
#hoplon
<
2017-08-17
>
thedavidmeister04:08:35

hmm, it appears that calling h/body anywhere in cljs tests triggers

thedavidmeister04:08:37

TypeError: null is not an object (evaluating 'this$.hoplonKids = null')

thedavidmeister05:08:54

woo, i'll update the other PR too

flyboarder05:08:36

im going to merge all my changes for 7.1 in the proposal branch

thedavidmeister05:08:58

yes, different PR

thedavidmeister05:08:13

@fiddlerwoaroof i'm just cleaning that fix up now

fiddlerwoaroof05:08:30

thanks, I guess I misread what was going on

thedavidmeister05:08:10

there were two PRs 🙂

alandipert14:08:52

thanks thedavidmeister flyboarder for swinging into action :-)

flyboarder20:08:37

I’m still pulling all the changes together

flyboarder20:08:30

@thedavidmeister can you poke at the 7.1-proposal branch and see why the new tests break?

thedavidmeister23:08:28

@flyboarder i've been using spec more since we last talked about it

flyboarder23:08:00

Wanna give the proposal branch a try?

thedavidmeister23:08:08

it doesn't generate tests for you though, although you can use it to generate test data

thedavidmeister23:08:20

but i think we probably want to write tests that are directly related to what the attributes are supposed to do

thedavidmeister23:08:56

like, toggle should actually toggle the element, not just accept data

flyboarder23:08:55

Yeah I think a combination of the tests you provided and spec should solve that, spec tests the data correctness and tests check the side effects

flyboarder23:08:47

I am confident in the 7.1 version except for a few of the specs, which don't account for destructuring yet

thedavidmeister23:08:05

cool, well i think that my tests/fix for the warnings in cljs 1.9.655+ should go in first

flyboarder23:08:56

Agreed, we can fix the 655+ stuff and then I think that's a good basis for the new functionality

flyboarder23:08:40

That gets tests working in 7.0.3 and cljs deps in 7.0.4

thedavidmeister23:08:53

i can also make sure there are tests in place for the changes to multi-arity functions

thedavidmeister23:08:03

that wasn't throwing warnings in 655 but i think it does in 660

flyboarder23:08:09

660 throws the variadic invoke thing, so we may want to solve both of those at once

flyboarder23:08:50

It's part of 7.1-proposal if you want to look at the code, basically we implement the correct signature and call a renamed function of the current logic

flyboarder23:08:23

Wasted lines of code personally but I guess there are reasons

thedavidmeister23:08:04

well, should we merge in 7.0.4?

thedavidmeister23:08:12

655 is more important than 660 imo

thedavidmeister23:08:35

655 has a ton of bug fixes in it relative to the earlier versions that have spec.alpha

flyboarder23:08:07

Well 660 is already in the proposal so moving stable to 655 makes sense for now too

thedavidmeister23:08:27

literally the only change in 660 is throwing those warnings 😛

flyboarder23:08:25

Yeah let's do 655

flyboarder23:08:42

We can keep bumping the proposal version prior to release

thedavidmeister23:08:44

using versions earlier than 655 was getting unusable for me