Re: HTTP/2 : problem with section 3.5 Connection setup magic

On 16/07/2013 4:51 a.m., Martin Thomson wrote:
> On 15 July 2013 06:01, Amos Jeffries <squid3@treenet.co.nz> wrote:
>> I think this particular section 3.5 needs a second explicit note to the
>> effect that the opaque blob should be adjusted for the draft-specific
>> identifier.
> The intent was always the precise opposite here.  That magic string
> isn't a string, it's just a sequence of octets.  It should NOT be
> adjusted.
>
> For one, we don't have any indication that it would be as effective if
> it were to be changed.

Ah. Thank you. Okay I will code up the HTTP/1.1 agent to expect those 
exact octets.

I still think there should be a second temporary note to clarify that 
intention for the drafts so long as the section 3.1 special note exists.

Amos

Received on Tuesday, 16 July 2013 12:25:24 UTC