- From: Amos Jeffries <squid3@treenet.co.nz>
- Date: Sat, 22 Nov 2014 05:20:40 +1300
- To: ietf-http-wg@w3.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 22/11/2014 4:20 a.m., Patrick McManus wrote: > On Fri, Nov 21, 2014 at 2:19 AM, Martin Thomson wrote: > >> On 20 November 2014 17:28, Mark Nottingham wrote: >>> <https://github.com/http2/http2-spec/pull/647> >>> >>> Any thoughts? >> >> I like the correlation text, that's a real issue. > > > agreed - the term origin might be stronger than 'site' as currently > used. > > >> I'm lukewarm on the value of the SETTINGS/PING thing. TCP >> window scaling relies on knowing RTT, so another way of measuring >> it in a >> > > h2 bits are e2e That is wrong. Any h2 intermediaries will likely be varying the RTT based on work being done for unrelated clients traffic. Stream windows may also be scaled down/up at each hop depending on the various hops capabilities. The way the h2 spec is written intermediaries have to do a *lot* of transcoding work to pass traffic around. Clients workloads will impact on each others behaviour in randomly varied amounts as things scale. The only thing e2e about the h2 transfer is per-stream window size. And all that tells is how many bytes the network can afford to buffer along the way. In reality *HTTPS* (HTTP over TLS) is where the e2e predictability problems start to occur in measurable amounts. Since clients traffic is distinctly segregated from each other and A-to-B encryption enforces large sections of the network to become measurable single-hops. Amos -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (MingW32) iQEcBAEBAgAGBQJUb2ZXAAoJELJo5wb/XPRjGIMIAJdnhRjCzAHnOoMv9JjZsCqM 4K1Lk/gr6ET5r3X8bYCObFI+KmCxbiYbqReO22m1t/d0sbrMDZyb4jVr0GZYzHrk mXxXLQVgCrvIcWieBSArhw5K5Duifhgz1f/MYG8ePXealgVjWBREY2GnNPjYLGjV xd7KhJ08l1JF5FDIskIRQPulT6R+BwYj7Bg1fW66at6+2MNRZjIBmWS83ooORimL S4quzqn3B1hk+zk4+QdsiUSUw9jDwfH4k6txWG7mQftGwGX6EFny67TiD5d3zh0n 0Le92TZvk4XA0NteU6ie9p00Er1ieSQiuHL21999hSbXZUt8bXo/Dc+rGkUXos4= =ML34 -----END PGP SIGNATURE-----
Received on Friday, 21 November 2014 16:21:27 UTC