Re: Questions about stream state transitions

(2013/07/18 17:08), Ilari Liusvaara wrote:
> On Thu, Jul 18, 2013 at 03:38:47PM +0900, Shigeki Ohtsu wrote:
>> Thanks for answers, they are very helpful for me.
>>
>> Following comments, I've update the table in
>>
>> http://html5.ohtsu.org/HTTP2_Stream_State.pdf
>>
>> Please review this if it is correct and I've also submitted a new PR of
>
> Eh, HEADERS in Reserved (remote) transitions to half closed (local)
> regardless of END_STREAM? That looks bit suspicious.

Yes, it can be an option that it is UNCHANGED to still remain in the reserved(remote)
when HEADERS without END_STREAM is received.

> Also, I think that anytime DATA can be successfully received,
> WINDOW_UPDATE can be sent (half closed (local) seems to violate
> that).

PRIORITY can also be received in reserved(local/remote) accrding to the above model.

I'm not sure that there are pros and cons about this model. If not, I think it would be a design choice.

Received on Thursday, 18 July 2013 08:33:26 UTC