[foms] Proposal: adaptive streaming using open codecs
zaheermerali at gmail.com
Wed Nov 3 11:50:22 PDT 2010
On Wed, Nov 3, 2010 at 6:03 PM, Andy Berkheimer
<andyberkheimer at youtube.com> wrote:
> On Wed, Nov 3, 2010 at 8:07 AM, Zaheer Merali <zaheermerali at gmail.com> wrote:
>> On Wed, Nov 3, 2010 at 6:06 AM, David R <videophool at hotmail.com> wrote:
>> > On Tue, 2 Nov 2010 14:19:38, Chris wrote:
>> >>We've actually got 3 use cases here, kind of mixed. "Standard" streaming,
>> >> ala youtube, adaptive, and also live.
>> > Yes, there are 3. The question is" are all 3 use cases covered by one spec?
>> > Live is also adaptive (i.e. Http Live Streaming), but with very different
>> > requirements, and system issues.
>> > David
>> Really 4. But only 2 where we need a solution for.
>> Ondemand non-adaptive (normal http can handle this fine)
>> Ondemand adaptive
>> Live non-adaptive (again normal http can handle this fine)
>> Live adaptive
> This is what I keep trying to get at - adaptivity is _not_ the only
> benefit of having greater client control over the stream with chunked
> requests, whether they be physically separate files or a single file
> with byte range requests.
> Greater control over buffer readahead and the ability to separately
> address segments of the video are also big benefits. I'd go so far as
> to say that those benefits are more important to us and can have more
> positive impact on the user experience than adaptivity.
Totally agreed. It is also beneficial server side for the chunked
requests as previously mentioned due to lack of custom streaming
servers required with just normal web servers required lowering the
barrier to entry and deployment costs.
More information about the foms