Re: Recovery from decompression failure (was: Re: Cost analysis: (was: Getting to Consensus: CONTINUATION-related issues))

On Wed, Jul 23, 2014 at 1:20 AM, Martin Thomson <martin.thomson@gmail.com>
wrote:

> On 22 July 2014 05:45, Ilari Liusvaara <ilari.liusvaara@elisanet.fi>
> wrote:
> > That's exactly what I proposed (I checked the spec, seems to be allowed
> > to send two settings in one frame).
>
>
> Yes, that is specifically allowed (we want to permit stateless
> processing of settings).
>
> I'll note that the change needs to be acknowledged in a header block
> with an encoding context update.
>
>
​If 2 SETTINGS_HEADER_TABLE_SIZE is in SETTINGS, say 0 and N in this order,
do we have to encode 2 encoding context update for both 0 and N or for just
N?​

​Best regards,
Tatsuhiro Tsujikawa​