5 ms·

Is there somewhere where you or someone else goes into more details on the points you made?

• jacquesm 12y ago
  PHK's post would be a good start: https://www.varnish-cache.org/docs/trunk/phk/http20.html https://www.varnish-cache.org/docs/trunk/phk/http20.html previous discussion on HN: https://news.ycombinator.com/item?id=6014976 https://news.ycombinator.com/item?id=6014976
  • youngtaff 12y ago
    PHK simultaneously complains that HTTP/2 doesn't go far enough e.g. replace cookies with a stateful mechanism, while also complaining the header compression and other mechanisms in HTTP/2 introduce the requirement for the server to maintain state. PHK also refers to HTTP/2 as a political protocol and casts various aspertions about those promoting it but you need to remember that PHK isn't a neutral observer - Varnish does't support TLS yet (though PHK is very clear on why this is the case).
    • Dylan16807 12y ago
      >PHK simultaneously complains that HTTP/2 doesn't go far enough e.g. replace cookies with a stateful mechanism, while also complaining the header compression and other mechanisms in HTTP/2 introduce the requirement for the server to maintain state. Sounds valid to me. Either retain state or don't.
      • youngtaff 12y ago
        Maybe but changing the state semantics on the client is more work than just changing the protocol stack. I suspect it will come in time but I'd argue it's better to develop incrementally rather than a big bang approach, and there's plenty in HTTP/2 to allow further improvement.