Editor-ready: Frame size (to address #553)

<https://github.com/http2/http2-spec/issues/553> is now marked as editor-ready, with consensus captured there.

I don't hear much controversy over 24 vs. 31 bits, so I'll leave it to the editor to come up with the exact format of the frame, based on the discussion. 

I do hear some pushback on the 256 byte minimum. Keeping in mind that the issue we're addressing is about recipient overhead in the other direction, as well as the potential impact on interop, I don't see consensus for such a low minimum. Let's keep the minimum at 16K for now; if anyone is sufficiently motivated, they can open a separate issue.

There was discussion a while back about not making this a granular setting (e.g., allowing any payload size between min and max to be set), and there seemed to be support for making it more coarse-grained (e.g., specifying the number of bits available in the field, so that the granularity is 2^n, leaving us with a setting of "14" to indicate 16K max frame size). I'll leave it to the editor to come up with a sensible scheme. 

If people have feedback about the choices the editor makes, I'm sure he'll be enchanted to hear about it.

Cheers,

P.S. This thread is drifting off into very exploratory / "we might need that" territory, and that's not good. Please stay focused on the actual issues, and remember that our charter is to replace HTTP for its common use cases, not to design a protocol for low-latency lightbulbs. Thanks.


On 12 Jul 2014, at 4:46 pm, Mark Nottingham <mnot@mnot.net> wrote:

> There has been a lot of discussion over the last two weeks about various proposals to address a number of issues. While we're not at the point where we have consensus to accept any of them wholesale, I do think we can reduce the surface area of the discussion by declaring consensus on the less controversial parts.
> 
> So: it appears that we have consensus to address issue #553 by:
> 
> * Expanding the frame size field to 24 bits
> * Reserving additional bits to align 
> * Adding a setting advertising the maximum frame size allowed by the recipient, with a default of 16K octets and a minimum of 256 octets
> 
> This would address (only) <https://github.com/http2/http2-spec/issues/553>. 
> 
> Does anyone have a problem with that, or further comments?
> 
> Regards,
> 
> --
> Mark Nottingham   https://www.mnot.net/
> 
> 
> 
> 
> 

--
Mark Nottingham   https://www.mnot.net/

Received on Tuesday, 15 July 2014 01:34:31 UTC