- From: Roberto Peon <grmocg@gmail.com>
- Date: Tue, 25 Feb 2014 16:44:29 -0800
- To: Martin Thomson <martin.thomson@gmail.com>
- Cc: HTTP Working Group <ietf-http-wg@w3.org>
Received on Wednesday, 26 February 2014 00:44:56 UTC
It implies that a security-conscious entity will want to keep the limit at some moderately large, but finite value, e.g. 1000. -=R On Tue, Feb 25, 2014 at 4:37 PM, Martin Thomson <martin.thomson@gmail.com>wrote: > https://github.com/http2/http2-spec/issues/416 > > Roberto, as the former strongest advocate for the current unlimited > initial value, does this extend to a limit on the stream count? Or is > it just a recommendation that we make in security considerations? > > Either way, I think that it's an easy thing to do. > > It reduces the number of guesses that can be made against HPACK. I'm > not sure if it goes far enough to address #373. Absent some > revelation, I'm of the mind that any shared context (browsers, > proxies) will have to restrict how items in the header table can be > accessed if we intend to keep HPACK. > >
Received on Wednesday, 26 February 2014 00:44:56 UTC