Re: [streams] [Reader] Should resolve readyPromise on getReader() call? (#258)

I am working on this + #260 in the reader-fixups branch. I was able to solve the waiting → waiting transition from https://github.com/whatwg/streams/issues/258#issuecomment-68828539 easily but the one from the OP is not as easy.

I'm interested in if you can make the alternative approach work somehow. Would appreciate hearing how you guys are doing on that. The design principle to keep in mind from the design doc is:

> Subclasses of ReadableStream should get locking support "for free." The same mechanisms for acquiring and using a lock should work flawlessly. More interestingly, if they wanted to support modifying the behavior of e.g. read() (or state or ready or closed), they should only have to override it in one location.

> Which location is more friendly? Probably in ReadableStream, so that ExclusiveStreamReader still works for ReadableStream subclasses. Less work.

(We have given up on "level 3" support, i.e. custom readable stream implementations. That seems fine. But subclassing is still important.)

---
Reply to this email directly or view it on GitHub:
https://github.com/whatwg/streams/issues/258#issuecomment-69679796