Adam Rice

This now sounds related to a general problem with `desiredSize`: often you don't want any extra buffering, but you do want to receive writes of some ideal size. This is...

See https://github.com/whatwg/encoding/issues/72#issuecomment-396882904 and so on for the historical background. I'd summarise it as: subclassing from TransformStream places constraints on implementations that don't benefit users. In the case of ReadableStream and...

My preference is to punt on this issue until we add a finally() method. In the meantime I think we can say that as with write(), handling errors in strategySize...

I find the "internal slot" style easier to read. With the "associated concept" style you end with sentences like > If this’s do not flush is false, then set this’s...

The goal of avoiding the queue bloat from TransformStream is a good one, but the details of how to do it are tricky. I think it _may_ be sufficient to...

> > WritableStreamDefaultControllerGetBackpressure(_controller_) performs the following steps: > > > > 1. Let _desiredSize_ be ! WritableStreamDefaultControllerGetDesiredSize(_controller_). > > 2. If _controller_.[[strategyHWM]] is 0, > > > > 1. Return...

I think I see two options for the API: 1. Implicit. Semantics are "`writer.ready` is true when the underlying sink is not writing". 2. Explicit. Semantics are "`writer.ready` is true...

See also #1025.

Does this mean that `response.unusable` would be equivalent to `response.bodyUsed || response.body.locked`?

Oh, sorry, I misunderstood. We've discourage the use of `disturbed` and AFAIK no-one outside of Fetch uses it.