Re: Set-Cookie and Cookie Optimized Binary Encoding

Part of that escape hatch ought to be a bit in the encoded header that
indicates a binary or text value. If it's text, then the 1.1 format is
used. If it's binary, the optimized format is used and we get a closest we
can get compromise on the translation from one to the other.
On Jan 19, 2013 5:28 PM, "Nico Williams" <nico@cryptonector.com> wrote:

> On Sat, Jan 19, 2013 at 6:11 PM, James M Snell <jasnell@gmail.com> wrote:
> > Yes, indeed. I'm wondering if there's a reasonable balance we can support
> > here... That is, provide for backwards compatibility for most but not all
> > 1.1 features and optimize for the most commonly used bits. Restricting
> set
> > cookie syntax could be one of those compromises.
>
> We should try to optimize bridging between HTTP/1.* and HTTP/2.0, but
> we should not penalize a pure HTTP/2.0 future in the process.  The
> obvious thing to do is to provide an escape hatch for contents we
> don't optimize and not try to optimize absolutely everything.
>
> Nico
> --
>