[opendtv] Re: NTIA: National Broadband Map has Helped Chart Broadband Evol
- From: Craig Birkmaier <craig@xxxxxxxxxxxxx>
- To: "opendtv@xxxxxxxxxxxxx" <opendtv@xxxxxxxxxxxxx>
- Date: Thu, 13 Aug 2015 09:58:40 -0400
On Aug 12, 2015, at 8:19 PM, Manfredi, Albert E
<albert.e.manfredi@xxxxxxxxxx> wrote:
Craig Birkmaier wrote:
One must question what would happen to those 25/3 speeds, if everyone
signed up tomorrow and started watching Netflix.
Which is not physically possible. These things take some amount of time.
What a flip flop. For months you've been saying the Internet is ready. Lately
you've been saying all would be well if the cabled MVPDs recovered all of the
bandwidth used for MPEG TS streams and just improved broadband.
Yes, it will take time and many more billions in investment. Meanwhile MVPD
services are delivering hundreds of HD streams to 80 million homes, accounting
for the majority of all TV viewing.
As we saw, though, Netflix was not averse to installing its own servers.
Parenthetically, I might be one of few households left in our neighborhood
WITHOUT FiOS or Cox. But the days of DSL are numbered, clearly. As I said,
Verizon is not repairing DSL, they say, if anything goes amiss.
Kilroy is correct. For video, edge servers are essential from an economic
perspective.
Why are you holding out on better broadband?
Craig, pay attention: I never implied that all TV traffic would have to go
through the backbone, quite the opposite.
Changing the subject again as usual...
The reality check is that the last mile cannot handle the shift from
broadcasting linear streams to millions of UDP streams.
Except you missed what Kilroy was getting at. The problem is the silly
proliferation of protocols.
That's part of the problem, but this has been the case for a very long time...
The problem is lack of agreement to use industry standards, too many DRM
standards, and rapidly evolving video codec technology. Compressed video
streaming is a hugely important market that attracted many flies.
The early MPEG standards were focused on high bandwidth/high quality streams -
MPEG-2 was co-opted for digital television broadcasting. MPEG-3 was to address
Internet streaming but was stillborn. The agreement for ISO MPEG and the ITU to
collaborate on h.264 was a major breakthrough.
But there were literally dozens of companies developing codecs or just creating
FUD. Microsoft was among the worst, starting with the DTV Team, then funding
the HD DVD Blu-Ray war, and trying to get SMPTE to adopt Microsoft's
proprietary compression technology. Sorensen, Duck, and many others drove
things on the proprietary side.
Flash has a storied history. It was much more than a video streaming codec.
I worked with the companies that created the precursors to Flash - the VC
community poured a bunch of money into "multimedia." Most of those bets were
consolidated into Macromedia, which helped to popularize "Director," an
application to created animated vector graphics for websites and CD-ROMs. Adobe
acquired Macromedia and integrated it into their web application suite as
Flash. Years later they worked with Duck to create the Flash video codecs that
you are so enamored of.
It is important to note that the major value of Flash was not video streaming,
but the creation of sophisticated websites using efficient vector graphics.
This was the BIG problem HTML 5 was created to solve. Even HTML 5 does not
dictate a video codec or streaming protocol, but new versions of HTTP and h.264
now dominate the web, as RTMP fades into oblivion
Fortunately getting past Flash has helped immensely. And we can thank Apple for
helping to move to a more standards based environment by refusing to support
Flash in iOS.
Flash was the first cross-platform web streaming solution.
Far from it. There were many over the years.
The others came after. So Flash was the lingua franca, and it was supported
even by Android initially.
Flash became the lingua Franca because it was already widely used to build
websites. Adobe was able to work with Duck to keep improving the video codec,
but Flash was well established with nearly two decades of development and
support, before video streaming started to take off around 2005, when You Tube
opened its doors using Flash. But there was another, far more important market
that really helped to establish Flash - porn.
Flash provided the structure to build sophisticated interactive websites, and
the ability to navigate video clips with all kinds of graphic support.
Android did support Flash initially, but dropped that support for the same
reasons that Apple stated when they announced they would not support Flash in
iOS. The biggest problem was power consumption when running the Flash video
codec on the main CPU. The nail in the coffin for Flash was the decision by GPU
vendors to support h.264 in their graphics chips.
The inability of Adobe to deliver an efficient Flash player for mobile devices
was a major problem as well, which led Adobe to throw in the towel and support
HTML 5.
Kilroy was lamenting the state of chaos, not supporting it. This is the quote
from Kilroy:
"Still, it won't be adequate for projected video growth **unless cache
efficiency is optimized by reducing the number of redundant encodings
required** ..."
That's what standards are for. But there is a ying and yang, as many companies
want to be among the encoding required.
Did you get that part that I emphasized? It *always* scales better when
everyone uses the same standard. Imagine how the Internet would work if this
weren't the case.
Like it did with Flash and many other technologies that proved there was a
market before the IETF pulled industry players together to codify the best
stuff into their standards.
The real magic of the Internet is that anyone can create a proprietary standard
for something and drive innovation. It is the codification of things that have
been proven in the real world that has made the Internet such a powerful
infrastructure for innovation.
Contrast that to something like the ATSC or DVB, which created proprietary
standards that have had relatively short half lives.
Imagine if every tom, dick, and harry used their own proprietary packet
headers and routing protocols. Same applies to codecs and streaming
protocols, and everything else. This IS NOT TO SAY that different streaming
protocols CANNOT be sent over the Internet, Craig. This is to say, **that
approach does not scale well**.
Nothing scales well forever. We are still transitioning to IPV6.
Codecs and related technologies will continue to evolve. When a new one becomes
popular you still have to support the legacy stuff until the devices are
upgraded to support the new standards. And unfortunately, we still live in a
world where powerful interests dictate the devices they want to support...
Such is the reality of a competitive marketplace.
Regards
Craig
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:
- Using the UNSUBSCRIBE command in your user configuration settings at
FreeLists.org
- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
unsubscribe in the subject line.
Other related posts: