W3C home > Mailing lists > Public > ietf-http-wg@w3.org > April to June 2013

Re: Out-of-order Frames

From: Mike Belshe <mike@belshe.com>
Date: Sun, 23 Jun 2013 15:53:05 -0700
Message-ID: <CABaLYCtui3FZE6BmjtyzbUSpRrH6M+mXS1cP7SkMxcG9ApsDMg@mail.gmail.com>
To: Sam Pullara <spullara@gmail.com>
Cc: Amos Jeffries <squid3@treenet.co.nz>, httpbis mailing list <ietf-http-wg@w3.org>
It seems to me that a combination of client-side templates and server push
could do this with HTTP/2.

The client side template is effectively the placeholder you speak of.  The
slow-to-compute server resources would be server push streams.  Javascript
would glue them together such that receipt of the server push would put the
content in place.

One thing we haven't defined (and maybe this is where WHATWG could help) is
with Javascript APIs for server push.  For instance, what if the server can
push a resource that the client never requests?  Some simple javascript
event notifications could really be interesting here.

Mike





On Sun, Jun 23, 2013 at 9:25 AM, Sam Pullara <spullara@gmail.com> wrote:

> The OBJECT, IFRAME and similar solutions don't work because they aren't in
> the same document.
>
> Sam
>
> On Jun 22, 2013, at 7:18 PM, Amos Jeffries <squid3@treenet.co.nz> wrote:
>
> > On 23/06/2013 1:45 p.m., Zhong Yu wrote:
> >> this reminds me another request for out-of-order delivery of an http
> entity
> >>
> >> http://lists.w3.org/Archives/Public/ietf-http-wg/2012OctDec/0259.html
> >>
> >> (it would be funny though to implement out-of-order frams on top of
> >> TCP which already deals with out-of-order segments)
> >
> > It does sound very similar. But the kfall draft is very close to
> technically possible.
> >
> > The case as presented by Sam the "chunk 3" starting offset is completely
> unknown to both client *and* server. Take an arbitrary length string,
> insert a second arbitrary string (length M) into it at a pre-known starting
> offset (N) and somehow determine the value of N+M before M comes into
> existence. Not possible until we fix our machines phychic abilities.
> >
> > This is a solved problem in markup languages, they use a special tag
> (OBJECT and IFRAME I referred to earlier, IMG, SCRIPT, VIDEO are others in
> HTML) in a tree structure which can be adjusted - you should all recognise
> the DOM. However outside of markup languages multimedia resources are not
> able to split their content over a tree-like structure without becoming
> something else.
> >
> > Amos
> >
> >>
> >> Zhong Yu
> >>
> >>
> >> On Sat, Jun 22, 2013 at 2:34 PM, Sam Pullara wrote:
> >>> Commonly in dynamically generated websites there are sections of
> content that are static and parts that are calculated on a per request
> basis. The current best practice for accelerating the delivery of a page
> like this involves leaving identifiable DOM elements where the dynamic
> content would appear, flushing the entire static page, and then flushing
> JavaScript script nodes as calculations complete (e.g. Facebook's BigPipe
> and deferred rendering in mustache.java). This practice only works for HTML
> pages (with JavaScript enabled) and offers no acceleration for other types
> of content delivered over HTTP.
> >>>
> >>> One possible solution to this problem would be to allow for
> out-of-order frames where the static frames are sent as quickly as the
> connection allows and dynamically generated frames are then sent later as
> they become available on the server. We would likely not want to enable
> this in general and would likely need to negotiate this behavior between
> client and server. Looking at the spec, frames might not be the right place
> but something on top of frames because of the size limitations.
> >>>
> >>> Has something like this been discussed before? Would this be the right
> mechanism or are there better ways to do it?
> >>>
> >>> Thanks,
> >>> Sam
> >>>
> >>>
> >
> >
>
>
>
Received on Sunday, 23 June 2013 22:53:33 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 11:11:13 UTC