- From: Cory Benfield <cory@lukasa.co.uk>
- Date: Tue, 6 May 2014 14:35:23 +0100
- To: Poul-Henning Kamp <phk@phk.freebsd.dk>
- Cc: Julian Reschke <julian.reschke@gmx.de>, Larry Masinter <masinter@adobe.com>, Yoav Nir <ynir.ietf@gmail.com>, "K.Morgan@iaea.org" <K.Morgan@iaea.org>, "ietf-http-wg@w3.org" <ietf-http-wg@w3.org>, "C.Brunhuber@iaea.org" <C.Brunhuber@iaea.org>
On 6 May 2014 13:50, Poul-Henning Kamp <phk@phk.freebsd.dk> wrote: > Or for that matter: If HTTP/2 will be an improvement in the first place. > > I've yet to see any numbers that justify technical cost of the added > complexity. I'm not quite as pessimistic as you are, Poul-Henning. I'm sure there will be some specific use-cases where HTTP/2 is a win, which is why I want client libraries in Python to have access to it. But I agree, it seems that there are going to be a reasonable number of use-cases where HTTP/2 is less effective than HTTP/1.1.
Received on Tuesday, 6 May 2014 13:35:55 UTC