Adam Rice

> maybe would be grate if they could be similar (isomorphic)? If I'm reading the doc correctly, it will be identical, except we probably won't need an options parameter.

One of the design principles we tried to follow was "users shouldn't be punished for race conditions outside of their control". We appear to have failed in this case. I...

I created #616 to discuss TransformStream byte streams independently of this particular pull request. Hopefully we can decide the direction there and then come back here.

Random observation: I find writableType/readableType confusing and would prefer to have inputType/outputType.

I didn't realise that this had already been bikeshedded. I would like to be pedantic and say that writableType/readableType are from the point of view of the author of the...

I think we've been pretending that resolving a Promise with a non-JavaScript type is a meaningful operation, and mostly getting away with it, but now we've run out of luck....

- size() does nothing for byte streams. Related: #729 - highWaterMark is copied so there's no way to make it change over time - I think size() has to throw...

I have a tentative plan for this: 1. Move tests that are relevant to both types to a new directory, readable-streams-bikeshed. "relevant" usually means that the test result doesn't depend...

So maybe something like * `resources/common/foo.js`: _definitions of tests for the foo feature_ * `readable-streams/common/foo.js`: _wrapper for `type: undefined`_ * `readable-byte-streams/common/foo.js`: _wrapper for `type: 'bytes'`_ ?

This would mean changing the calculation of desiredSize from being directly based on the queue to being tracked separately. This seems reasonably achievable, but less elegant. I'm concerned that callers...