mcnally
|
|
response 1 of 290:
| 
Jan 19 05:54 UTC 2006 |
In Item #36, jep wrote:
> I would be fascinated to hear more about homes having a lot more network
> devices.
Thirty years ago (1976), home computers as we know them now were only barely
beginning to exist and probably nobody you knew had one. (The Apple II,
for instance, didn't come out until 1977..)
Twenty years ago (1986), computers were still largely the province of
enthusiasts and businesses but everybody (in modern, industrialized nations)
knew someone who had one and it seemed like everyone who didn't have one
was thinking of buying one.
Ten years ago (1996) more or less every affluent and most middle class homes
had a computer and many people were hearing about something called "the
Internet" for the very first time.
Now it's 2006 and computers have become part of the daily life of everyone
who is reading this and probably most of the people they know. Many households
have multiple computers. What's more, powerful computing and storage devices
have worked their way into the personal effects that people are accustomed
to carry around with them -- cell phones, MP3 players, PDAs, etc..
Pretty much everyone agrees, computing is becoming ubiquitous and as this
happens many of our devices are starting to talk to each other.
John asks what kinds of things we can expect to see connecting to each
other via computer networks in the near future. Let's start with the obvious:
computers and computer peripherals. Many families are multi-computer now
and some are approaching one (or even more than one!) computer per member
of the family. In some places kids are bringing home laptops that are
provided by their schools. And In families that have high speed / broadband
internet access people are finding ways to share their connections between
multiple computers. They may also be sharing other things, which may or
may not live directly on the network. For example, while it may make sense
for many families to have more than one computer, not too many families (yet?)
have a need for more than one printer -- most share. Some share by connecting
the printer to a networked computer and having the computer take care of
spooling print jobs but it's also quite easy to buy a network-aware printer
that can connect directly. Expect to see similar arrangements with things
like digital (still) camera docks, digital video cameras, hard drives
(check out the rise of home NAS (network-attached storage) appliances like
the Linksys NSLU2) and all kinds of other things. Peripherals are moving
off the computer and taking on a semi-independent existence on the network.
All the while this is happening there's another revolution going on, too,
in digital entertainment. Technology pundits have been predicting digital
hubs in the living room for years now but we're really getting there.
Depending on your tastes and those of your family members, some network
devices you might have in your living room include: PVR devices such as
TiVo and ReplayTV, videogame consoles such as the PS2 or XBOX (regular or 360),
media appliances which stream your music (and increasingly, your videos)
from your computer's hard drive to your living room stereo, and more.
What about other rooms in the home? Well, a lot of families have televisions
and telephones in family rooms and bedrooms. You might not know it yet but
the PSTN (public switched telephone network) gives every appearance that it's
dying. It was seriously wounded by affordable cell phone service and VoIP
telephony stands ready to deliver the coup de grace. Very soon now your
telephones will either be network attached devices or you'll have an
adapter which sits on the network and mimics POTS over your home's phone
wiring. Television is changing, too. There's a collossal battle brewing
between phone companies, cable companies, and ISPs and by the time it's
over you will more than likely be buying what the industry calls "the
triple play" (voice, video, and data services) from a single provider.
(You might have more than once choice of providers, but the idea is you'll
buy them all from the same company and they'll come into your house on the
same wires.) VoIP phone service and television over IP video service are
already here in my home in Ketchikan -- how long could it possibly take
before they're available through most of the rest of the country?
too.
|
mcnally
|
|
response 2 of 290:
|
Jan 19 06:23 UTC 2006 |
A little bit more about the services I'm buying from my employer,
the local phone company..
Like pretty much every other house in the USA and Canada, I've got
a phone line connected to my house. Except in my case, instead of
getting a POTS service over that line, I get no POTS service, just
high speed ADSL2+. The speed we can get over that link varies
according to the quality of the wiring and distance from the central
office but for most customers we can give them 15-20 Mbps over their
regular copper phone wire. We partition that network connectivity
into several different virtual circuits which are assigned to VLAN
groups -- on a per household basis there's one virtual circuit that
carries video data over IP, one that carries phone calls over IP,
and one that carries data traffic to/from the internet. Reliable video
requires a huge amount of bandwidth but with 15-20Mbps to work with
we can easily provide three 3-4 Mbps video streams and still have
plenty of bandwidth left over for phone calls and, say, a 1 Mbps
data service. Or, if the customer doesn't have a need for three
simultaneous video streams (e.g. TVs in three different rooms, or
two TVs and one video recorder, etc..) we can give them more data
bandwidth..
In my house, then, I've got the copper phone pair coming from the
phone company to the outside of the house. It's hooked into a
big NID (Network Interface Device) which decodes the ADSL2+ signal
and separates it into the three LAN groups. The telephone wiring
inside my house is hooked into this NID and the NID also has an
analog telephone adapter built into it. If I pick up my phone I
get a dialtone, but not from the phone switch at the phone company,
it comes from the NID on the outside of my house and when my phone
conversation leaves my property it does so as IP packets flowing
over the high-speed DSL network. The other two LAN groups are
carried over twisted pair cabling to jacks in my living room, where
I've got a wireless access point / network switch combo box plugged
into the data jack so I can share the data connection among multiple
computers and I've got twisted pair that runs to the set-top box
attached to the (so far) single television that's using the service.
In most respects the television service isn't that different from
a regular digital cable service but there are some cool things we
can do with it. Unlike a regular digital cable service, which is
generally all multicast (every channel gets streamed to every
cable box that's allowed to receive it, all the time..) our network
can also do some cool stuff with unicast which makes for some
interesting potential with video-on-demand. Unlike a cable company's
pay-per-view, our video-on-demand isn't multicast, it's unicast
straight to your set-top box and the box can talk back to the server
that's streaming it the content. That gives us the ability to do
some things pay-per-view can't, such as start a movie any time the
customer wants it (not just when it's scheduled and pause, rewind,
or fast-forward the movie that's streaming from the video on demand
service. In about a month, when our full service launches, our
customers should be able to pick from about 6,000 titles, any of
which they can rent (for 24 hours) and start watching at any time.
I'm quite optimistic that it's going to beat going to the video
store in the rain.. Intriguingly, since we control the video-on-demand
service we'll have the ability to offer free local content. Want to
watch the high-school football game or tune into last night's town
council meeting? Well, maybe not, actually, but the point is we'll
be able to provide community oriented programming on demand.
Now consider -- we're a small company on a remote island and this is
the first, or maybe second generation of this type of service.
What will your (much larger) provider be offering you 5 years from now?
|
marcvh
|
|
response 4 of 290:
|
Jan 19 07:02 UTC 2006 |
I can't speak for your setup, but here in Seattle, Comcast's Video On
Demand service is multicast used in an attempt to simulate unicast.
There's a block of channels (I believe it's about 40) set aside for VOD,
and when you select a program it allocates one of those channels to you
and runs your program on it. You share those channels with all other
subscribers on your "neighborhood node" which has something on the order
of 400 of your neighbors.
This means that if more than 10% of Comcast customers are using VOD at
once, it fails. That's not necessarily a big deal, most utilities are
allocated that way, and it's always going to be possible to scale the
service by making more nodes which each service a smaller customer base.
But it also means that if you have a QAM-256 tuner (which many HDTVs do
these days) you can just flip through the VOD channels and see what your
neighbors are watching. You can't tell which neighbor it is, of course,
and if the neighbor pauses or fast-forwards the content then you'll see
that right along with him. I believe this works even for premium VOD
content though I haven't really tested it.
Unfortunately, they're still struggling to figure out how to make VOD
work well. The UI at the moment is crappy, and the navigation options
are very limited (you can fast-forward and rewind but only at once
speed, there's no chapter settings or menus or bonus content.) Very
little of the content is HD, and a lot of it is pan-and-scan (ugh.)
Most people are reluctant to pay $4 to watch a movie on VOD when the
DVD becomes available for rental or purchase weeks earlier and gives
you more for less. But I understand why the cable companies are
pursuing it -- it's one service which DBS would have a very hard time
doing given the nature of the technology. But I still predict that VOD
will remain a niche product, not due to limitations of the technology
but due to business decisions.
|