Stephen Hicks

Presumably one could take advantage of the (relatively) new template literal types to generate many possible assertions with only a few lines of declarations. I'd love to be able to...

Is there a compelling reason to invent the new syntax of having `import` be indexable and callable? At the very least, this makes it much more difficult to polyfill, and...

My understanding is that there are polyfill options for other browsers, making appropriate typings still pretty important.

Either way, this needs to take advantage of the new differently-typed getter/setter feature in order to be properly backwards-compatible. For example, ``` interface Element { get outerHTML(): string; set outerHTML(html:...

This looks much better. IIUC, the TypeScript team has a corpus of real world code that they can compile against - presumably they'd want to test this against that corpus...

Context on the recently-linked #21152: this is about a `readonly interface` keyword, which might allow a way to fix this without breaking existing usages. Basically, if a `readonly interface` had...

Not a maintainer, but I'll add that this is particularly relevant in light of the [Records and Tuples proposal](https://github.com/tc39/proposal-record-tuple) that's progressing through TC39 (#49243 already references this issue). It would...

Furthermore, if I move `ArrayList` to its own file, rather than nesting it, I also end up with a very strange `AbstractList.able` class, which is marked as yet another alias...

Closure Compiler is working on plumbing ES6 all the way through the compiler. Once that's done, constructors will no longer necessarily be conflated with the class itself, so it could...

Not directly. They're related in that passes that depend on type information will probably want to only migrate in NTI. But in terms of usability for dossier, it should be...