Lasse R.H. Nielsen

I'd like to launch enhanced parts as soon as possible, and not delay it on augmentations. Then they should be different experiments

It's tricky. All we can safely say that the method does not appear to have a body. If it's inside a non-abstract class or will _likely_ get a body, ......

Using `external` is almost certainly wrong. It may work, but it's sending all the wrong signals, and even when it works, it's not better than `=> throw Unimplemented("Not augmented");` in...

> we should make an official distinction between syntax errors and semantic errors? Yes. We should define when something is an error for a single syntactic declaration, and when it's...

That's what I expected. It does mean that `var x;` cannot be augmented to have a non-`dynamic` type by providing an initializer. Which I guess i fine, you should type...

It sounds like you're describing a general entity/relation model specification, with entities, properties and relations, and then the actual physical serialization format of that model is an implementation detail. Using...

Then I'd define the extension types first, and the JSON schema to match the API, not the other way around. (But I'd also prefer to not expose the parsed JSON...

> depends rather on whether you consider the capability to cast to `Map` as exposing the format I do, unless we explicitly document that you should not depend on the...

We're not providing a Google service. We're not providing a service at all. We are providing a way to interoperate with the macro host of the current SDK tool that...

> That's simply not true, however much we would like it to be true. We have already broken google3 and Flutter via the macro experiment as a result. Are we...