Re: [whatwg/streams] Add a proof of concept of optimized pipe (#511)

> we would add an unobservable brand check and fast path

I see. Anyway we need to perform some special interaction with the streams to realize the optimization. Even the propagation idea is requiring the streams to expose special signal / method.

One good thing of leaving the decision to the streams and have the global manager to fetch some predefined signal from them as done in the propagation idea above is that each transform stream can control the timing to announce longer piping. E.g. one could consume some initial header to do something special, and then pass through data to the next streams.

> Maybe it should say both.

Yeah! I'm not yet sure about necessity of different backpressure signals. Let's just try.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/whatwg/streams/pull/511#issuecomment-245175906