- From: Steven Robertson <strobe@google.com>
- Date: Mon, 13 Jan 2014 12:50:16 -0800
- To: Anne van Kesteren <annevk@annevk.nl>
- Cc: WebAppSec WG <public-webappsec@w3.org>
Received on Monday, 13 January 2014 20:51:24 UTC
On Mon, Jan 13, 2014 at 12:18 PM, Anne van Kesteren <annevk@annevk.nl>wrote: > On Mon, Jan 13, 2014 at 7:52 PM, Steven Robertson <strobe@google.com> > wrote: > > The application I work on is latency-sensitive and uses the Media Source > API > > in concert with XHR to download media bytes. In order to avoid the > latency > > cost of a preflight to each host in our CDN, we have developed a > workaround > > where we use a 'range=' query arg to subset an existing resource. This > > workaround is suboptimal, and we would prefer to use a Range header, but > the > > latency impact of the extra request has a significant impact on quality > of > > experience. > > > > What are your thoughts regarding the addition of 'Range' as a simple > request > > header to allow for this use-case? > > So in this scenario you set the Range header yourself using XMLHttpRequest? > Correct. When the Media Source API is available, we parse the media index at the JS level and use that to manage media loading by making ranged requests from the app. This allows a single resource to be shared between platforms where we can co-optimize the application and CDN and those where we have no control over media loading. Thanks, Steve
Received on Monday, 13 January 2014 20:51:24 UTC