Observations of an Internet Middleman (Level3) (was: RIP

Owen DeLong owen at delong.com
Sun May 18 17:24:48 UTC 2014


On May 16, 2014, at 10:06 AM, Scott Helms <khelms at zcorum.com> wrote:

> Blake,
> 
> I might agree with your premise if weren't for a couple of items.
> 
> 1)  Very few consumers are walking around with a HD or 4K camera today.

Not true. Most cell phones have HD cameras. Most CCD video cameras sold in
the last 5 years are HD capable.

> 2)  Most consumers who want to share video wouldn't know how to host it
> themselves, which isn't an insurmountable issue but is a big barrier to
> entry especially given the number of NAT'ed connections.  I think this is
> much more of a problem than available bandwidth.

Yes, but NAT is a temporary problem that is already gone for ~40% of the
subscribers on the largest CMTS networks in the US and is disappearing for
everyone else fairly quickly.  It’s disappearing even faster for anyone who
buys an X-Box One or gets an IPv6 Tunnel.

An application on an X-Box One could literally solve the video hosting problem
overnight.

This is an example of the limitations on innovation posed by NAT which is one
of the reasons it’s becoming more and more important to move forward with
IPv6. Since there are enough drivers and that transition is already in progress,
treating it like it’s a bigger problem than available bandwidth really doesn’t
make sense to me.

Available bandwidth is the much more insurmountable barrier at this point.

> 3)  Most consumers who want to share videos seem to be satisfied with
> sharing via one of the cloud services whether that be YouTube (which was
> created originally for that use), Vimeo, or one of the other legions of
> services like DropBox.

Sure, but there are other more interactive services that are under greater and
greater demand and realistically, people will come to expect multi-party HD
video chat as a given over time. The reason they accept it not working so far
is because they haven’t seen it actually working. As it becomes more ubiquitous
in other parts of the world, demand will grow in the US.

Shared gaming experiences will be another driver. While games are engineered
today to deal with the limited bandwidth available, developers are seeking ways
to deliver a richer, more immersive interactive experience and that’s going to
require more bandwidth. Once NAT can no longer be blamed as the primary
barrier, bandwidth will be their next target.

> 4)  Finally, upstream bandwidth has increased on many/most operators.  I
> just ran the FCC's speedtest (mLab not Ookla) and got 22 mbps on my
> residential cable internet service.  I subscribe to one of the major MSOs
> for a normal residential package.

Good for you. I’m paying for business service at the middle tier in my area
and get 27Mbps down and only 7Mbps up, both in what my provider tells
me they are selling me and in most of my mLab _AND_ Ookla tests.

If I went with DSL, the most I could get would be 1.5Mbps down and only
384Kbps up.

I’ve been getting those same levels of service for more than 5 years now.

Upstream bandwidth is definitely a limitation and it definitely hasn’t improved
for many customers.

Owen



More information about the NANOG mailing list