Re: [quicwg/base-drafts] create codec streams only when necessary (#2090)

[Luca Niccolini <[e-mail]>]

@dtikhonov yes with the current text, resetting any control stream is illegal. 
However given that we are discussing about making qpack streams kind of special, we _could_ allow resetting them. And that is what I proposed in my first comment.

> Could we instead relax the constraint that qpack streams (being control streams) cannot be closed, and allow an endpoint that doesn’t want to do qpack to not open its control streams and to immediately reset (with a specific error code, yeah I know YAEC) the peer initiated qpack streams?

My original point about the complexity of implementation for little benefit seems to be same that others have, and leaving aside concurrency and implementation specific concerns (which I don't deny exists in our code, but they are easily fixed) my point was more general as I feel like we are trying to bake into the protocol an optimization that in my opinion is better handled in the implementation of the memory constrained devices and we should just make it possible, rather than adding more constraints to the common case. So generally speaking I feel more comfortable relaxing constraints rather than tightening them, if possible and if that does not pose a threat.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/quicwg/base-drafts/pull/2090#issuecomment-448530612