James Garbutt

it may be better to not imply a shadow root, i.e.: ```html ``` or some such thing. otherwise it seems we'd be unable to register an element which has no...

Seeing as we are coming to the end of 2019 (the year this repo was supposedly being deprecated in), now could be a good time to add the deprecated flag...

@beenotung TSLint has always been limited compared to ESLint, there are a whole bunch of rules it never had which ESLint did. Not to mention the plugins and much larger...

i was poking around this issue... ultimately its because the compiler's result ends up like this: ``` // eslint-disable-next-line no-undefconst props = { ...Astro.props } as Props ``` it then...

Except your current logic means it doesn't close when disconnected from dom. It explicitly tells it not to flip the `open` flag in the case, so when it gets reconnected,...

yeah im not sure its so simple.. we use sinon's own assertions too in tests so running in esm wouldn't be able to resolve that afaik the test only needs...

i've rebased onto master and done the suggested changes. the tests seem like something that, while it works in sinon, needs thinking about properly. so im not sure if we...

its a weird one, because even node's esm implementation isn't the same as how browsers work. they do a lot of funky stuff to make commonjs interop work. ideally we'd...

yup i know, ill have a play around with it today if i can. a test which just evaluates a script to test the exports exist in browsers would do...

i added your test... it works but to do the puppeteer side of it gets sketchy really quickly. basically, the thing we produce at the min isn't such a valid...