Robert Fletcher

Ah, yeah. Hmm... Well I was using `.to_i` to get past any time comparison issues, but maybe I'll just try removing that from my specs and see how it goes....

Yeah, we generally don't use `context` if there's only a single test for a condition.

I actually disagree with the quote: > In general, we recommend you use the loosest matcher that still specifies the behaviour you care about. I think it's best to specify...

I think it would break a *lot* of tests if RSpec were to change it. Maybe they could add a configuration for `strict_predicates` or something.

FWIW, we've defined an `invoke` matcher that captures this as well: ```rb expect { a[1] } .to invoke(:[]).on(a).with(1).and_return(2) ``` Though, like you say, we wouldn't normally test both the return...

For what it's worth, `isolatedModules=true` didn't solve the problem for us, nor did setting `--maxWorkers=1`. We haven't been able to upgrade past version 23 because our build times out due...

I was seeing this when the modal is unmounted without having been triggered. Adding this fixed the issue: ```js Modal.setAppElement('#app-base'); ```

@sksundram right now I've got it in each of the files that I'm importing `Modal` for rendering. I suspect you could put it just about anywhere, though, like in your...

Would it be possible to get this working using [`xvfb` on CI](https://docs.travis-ci.com/user/gui-and-headless-browsers/#using-xvfb-to-run-tests-that-require-a-gui)? It does seem like it should throw a helpful error, though, when an XWindow system is not available....

When I try `Image#capture` in a Rails console on my local machine it just hangs. At any rate, it does seem like we should detect whether there is an XWindow...