Andrew Duthie

I put a lot of value in the CSP, so I'd rather see it enforced in more places than in fewer (e.g. local development). One reason is that local development...

Thanks for the feedback, @dstaley and @developit ! To clarify, would the suggestion be to treat this similarly to the current "View Unchanged" collapsing ([example](https://github.com/WordPress/gutenberg/pull/21017#issuecomment-601193735))? Optionally as a separate (or...

Sorry y'all for the delay. I've been meaning to revisit this one. I like the idea to consider threshold as treated similar to what exists already with the behavior of...

Hey @DerekTBrown , I admit I totally forgot about this one. If you're keen to see it over the finish line, I'm more than happy to let you take the...

This is not yet (reasonably) possible. Related issue at https://github.com/WordPress/gutenberg/issues/2474 with some exploration of plans.

>What if we have blocks returning an array of elements? Does this mean we won't be able to parse attributes from these nodes (except the first one?) Actually, you're right,...

>I see some logic in having `undefined` for the first I could see this, but it's more unexpected in that `attr` only returns `undefined` here because the `body` wrapper we...

The other option for consistency is to always ensure a wrapper, even in the element signature. ```js const markup = ''; const element = document.createElement( 'div' ); // Before: parse(...

I still think the original proposal here is most sensible and makes for easy parsing of a root node, but acknowledge it causes some difficulty for multi-element blocks in Gutenberg....

Another thing I don't particularly like about forced selectors is it results in a choice between redundancy and inconsistency with regards to the `query` matcher. The behavior of `query` is...