W3C home > Mailing lists > Public > www-html@w3.org > November 2006

Re: Suggestion: http request bundle

From: Mark Birbeck <mark.birbeck@x-port.net>
Date: Fri, 3 Nov 2006 16:23:21 +0000
Message-ID: <640dd5060611030823t3cbf575boeb81191d77329b49@mail.gmail.com>
To: www-html@w3.org


Interesting idea. It may cross over with some of the web app packaging
work that is being mooted.

I don't know if you are aware, but another factor that makes your
technique desirable is that it seems browsers are only supposed to
send 2 requests at a time to HTTP servers. RFC 2616, section 8.1.4
says this:

  "Clients that use persistent connections SHOULD limit the number of
   simultaneous connections that they maintain to a given server. A
   single-user client SHOULD NOT maintain more than 2 connections with
   any server or proxy."

which, implies that each subsequent request by the browser will be
blocked until one of the two that are 'in progress' has completed.
Looking at the way browsers behave in relation to this, file size
seems to be a far less significant factor on performance than the
number of files referred to.



On 03/11/06, Doug Whitehead <Doug.Whitehead@nasd.com> wrote:
> Enhance the http protocol to create a request bundle, with a zip response
> to
> minimize the chatter between the browser and a server.
> > From the web user's perspective website load times could be accelerated
> > if N
> requests for resources could be bundled into a single request from a
> webserver.
> This would also minimize the load on the webserver, as one request bundle
> and zip response could replace what currently takes dozens of
> request/response pairs.
> Comment:
> Think about the dozens of gifs, icons, etc. that get loaded when one first
> hits a website. Seconds could be shaved off the initial load time if the
> browser were able to grab multiple items from its request queue for the
> same
> destination, same user, etc. To do so would surely suggest a minor
> Javascript API enhancement to take advantage of the request bundle
> efficiency.
> Migration Thoughts:
> As a request bundle is functionally equivalent to N individual request, a
> browser that encounters a site that does not implement the request bundle,
> could fail gracefully back to individual requests.
> > From the webserver's perspective, if the browser is incapable of issuing
> > a
> request bundle, the webserver will merely receive individual requests and
> thus issue individual responses, as is done today.
> Personal Comment:
> I am working with a team using AJAX. As a result of creating a rich client
> experience, at one point there are some 60 static resources, which can add
> 5
> seconds to the page load time. (icons for toolbar buttons, etc.) This type
> of experience can not be uncommon these days.
> Doug Whitehead
> NASD Architect
> Doug.Whitehead@nasd.com
> This email, including attachments, may include confidential and/or
> proprietary information, and may be used only by the person or entity
> to which it is addressed.  If the reader of this email is not the
> intended recipient or his or her authorized agent, the reader is
> hereby notified that any dissemination, distribution or copying of this
> email is prohibited. If you have received this email in error,
> please notify the sender by replying to this message and delete this
> email immediately.

Mark Birbeck
x-port.net Ltd.

e: Mark.Birbeck@x-port.net
t: +44 (0) 20 7689 9232
w: http://www.formsPlayer.com/
b: http://internet-apps.blogspot.com/

Download our XForms processor from
Received on Friday, 3 November 2006 16:23:38 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 30 April 2020 16:21:01 UTC