- From: Mark Nottingham <mnot@mnot.net>
- Date: Wed, 24 Oct 2012 21:12:12 +1100
- To: Amos Jeffries <squid3@treenet.co.nz>
- Cc: ietf-http-wg@w3.org
On 24/10/2012, at 7:40 PM, Amos Jeffries <squid3@treenet.co.nz> wrote: >> I'd like to validate an assumption that has been discussed in-person in a meeting (forget which one) but IIRC not yet on-list. >> >> It's that if we can design an upgrade that fails fast and reliably for HTTP URLs, it's acceptable for there to be a (say) 70% success rate initially, with the idea that it will rise over time -- perhaps slowly, in terms of browser releases, but relatively quickly, in terms of the lifetime of HTTP. >> >> Can we (everyone) agree upon that? > > > I thought we are working from a baseline of Upgrade: being *the* existing defined HTTP/1.1 method for negotiation. Those Chromium experimentd proving it had ~67% success rate. Now just looking for ways to use it more efficiently and alternative methods that would improve that success rate faster? Yes, that appears to be the case right now. -- Mark Nottingham http://www.mnot.net/
Received on Wednesday, 24 October 2012 10:12:23 UTC