- From: Roberto Peon <grmocg@gmail.com>
- Date: Mon, 21 Jan 2013 20:22:41 -0800
- To: Mark Nottingham <mnot@mnot.net>
- Cc: HTTP Working Group <ietf-http-wg@w3.org>
- Message-ID: <CAP+FsNfxhML_tiQo4ML2jE704O5fkwYuMDB4Jy9wv7rW_94bwA@mail.gmail.com>
Many mobile sites do content inlining because it REALLY makes a difference in latency/responsiveness. The problem of when to do this is the same problem. Those who do not do it have a clientele which suffers significantly higher latency. The mechanism to decide whether to do this or not can be based upon user-agent, dst IP, src IP, URL, or configuration of the server/proxy/accelerator, and is mostly orthogonal to the features of the protocol-- it is ultimately up to whomever is using the server, configuring the router/DNS, creating the content to decide how best to serve their content to the various devices. It is (I think) our problem to ensure that they have mechanisms available that allow them to do this in efficient ways when they so choose. -=R On Mon, Jan 21, 2013 at 8:06 PM, Mark Nottingham <mnot@mnot.net> wrote: > > On 22/01/2013, at 3:01 PM, Roberto Peon <grmocg@gmail.com> wrote: > > > The same way one would in a non-push world-- determine if it is mobile > and serve the most appropriate content... :) > > Yes and no. Today, you have the option of doing it server-side *or* > client-side. When the server is pushing, all the decisions are made > server-side. That means, for example, that it'll be potentially more > difficult to server a responsive site that's completely static. > > > -- > Mark Nottingham http://www.mnot.net/ > > > >
Received on Tuesday, 22 January 2013 04:23:09 UTC