Re: [quicwg/base-drafts] Consider a new stream type to act like MAX_PUSH_ID (#2418)

[Kazuho Oku <[e-mail]>]

> Consider defining a new unidirectional stream type for Max Push ID updates. This can be expressed as unframed data: a series of increasing var ints.

Am I correct in assuming that the design have the same issue to what you tried to address in #2412? If so, I wonder what the benefit of changing the design is.

> The use of MAX_PUSH_ID frame is entirely unidirectional, so using a uni stream is a slightly better fit to the concept. Another reason to consider this approach is that it lends itself to extracting server push into an extension. It provides a pattern for how future extensions could manage themselves without needing to pollute the core control stream.

I am not sure if that is the preferred approach. Although I acknowledge some of the upsides, the downsides of using unidirectional stream are that it requires more state (due to the need for it's own send / receive buffers) and that it is sometimes complex to synchronize the states that are communicated over multiple streams. There are multiple cases where we could use a dedicated stream instead of sending a frame on the control stream; PRIORITY, CANCEL_PUSH, GOAWAY, DUPLICATE_PUSH, I believe that the payload of all these frames can be sent using their own streams, the same way the proposal tries to carry the payload of MAX_PUSH_ID frame.

To put it another way, it seems to me that our current design pattern is to use the control stream for transmitting various data instead of allocating a different unidirectional stream for each type of information. Allocating a different stream for just MAX_PUSH_ID would be a inconsistency in the design. Do we want to have dedicated unidirectional streams for each type of control information? To me that sounds like a design that's unnecessarily heavyweight.

-- 
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/issues/2418#issuecomment-460761115