No local storage? If a stream is coming into a system, that stream can be recorded, no?

Murali,

You could probably easily record the stream.

However, if you live in the US, you would likely run afoul of the Digital Millennium Copyright Act (DMCA).

Basically, the DMCA makes it illegal to unscramble or decode any sort of encryption, no matter how trivial.

I doubt that the stream from the Cringemaster's office represents wide realworld experience. I have modest bandwidth DSL and with all the video crap that's going, streaming and downloads, I can no longer watch videos--too much stalling for buffering--, and it takes a loooong time to torrent. Many sites now load like they once did on dialup.

fuggedaboudid

This scheme will require a lot of upstream bandwidth in the access network. It will be of limited use for people on current cable and DSL connections, but it should fare better in FTTP environments. Also, the degree to which is offloads the carrier network by using local peers will depend on how close to the home user the carrier routes between them, it may be upstream of the worst congestion point, although it certainly varies across each implementation.

If the ISPs embraced this technology (like they should) and optimize their networks to use that intranet bandwidth Bob is always reminding us about, this technology could be quite a win. If your peers are all only a hop or two away, you could probably stream video at a bitrate that is much higher than what the ISP is nominally selling you.

So we'll see how reality decides to disappoint.

The fact that Neokast is not aimed at on-demand video poses a slight conundrum to me - IPTV/bittorrent has value to me for three reasons. First, I can watch the downloads at my leisure. Second, I can get access to content that isn't shown locally. A distant third, it's cheap. Neokast will be cheap but will the content providers let second-rate citizens like me (i.e. non-US residents) watch their streams? And how will I organize my life around their channels when prime time is at 2 in the afternoon? I suppose the answer might be a PVR for Neokast streams to let me time-shift, but now this is starting to sound like a problem I already have.

If the ISPs embraced this technology (like they should) and optimize their networks to use that intranet bandwidth Bob is always reminding us about, this technology could be quite a win. If your peers are all only a hop or two away, you could probably stream video at a bitrate that is much higher than what the ISP is nominally selling you.

So we'll see how reality decides to disappoint.

The fact that Neokast is not aimed at on-demand video poses a slight conundrum to me - IPTV/bittorrent has value to me for three reasons. First, I can watch the downloads at my leisure. Second, I can get access to content that isn't shown locally. A distant third, it's cheap. Neokast will be cheap but will the content providers let second-rate citizens like me (i.e. non-US residents) watch their streams? And how will I organize my life around their channels when prime time is at 2 in the afternoon? I suppose the answer might be a PVR for Neokast streams to let me time-shift, but now this is starting to sound like a problem I already have.

Ok, this is probably the most interesting thing I've read in the last 6 months. I've been totally fascinated by the concept of where television is going and how the internet is (and will be) affecting it's broadcasting and usage.

Come on, this is not really a breakthrough. P2P TV has been out there for at least 1-2 years.

There is another company doing something very similiar - Network Foundation Techologies in Ruston, LA. In fact, they got a $100K grant from the National Science Foundation to develop it.
nft-tv.com

Coljac hits the nail on the head - 'at my leisure'.
So running off to the bathroom is still going to be a problem with this new format...
I guess an answer to that question would be to insert advert breaks - yay, can't wait...

I'll stick to P2P :)

SplashCast offers something different. To those who would become netcasters it's the most interesting concept I've seen lately.

You could still go to the rest room. There is no reason the client couldn't allow pausing. It would just save the data to your hard drive. No need for Tivo since you are watching it on a computer with tons and tons of storage (most likely anyway).

"This makes video rights management much easier and ought to appeal to TV networks and movie studios."

Umm...what about users? Why would that want this?

Interestingly enough I have been watching Nerds 2.0.1 and original Nerds in the last few days. At the end of Nerds you comment that we should tune in in 10 yrs to see what happens. It is about time to revisit.

In Nerds 2.0.1 it is interesting how many companies are gone or a shadow of their former selves and it seems the only way to survive is to start companies and move on, christine comaford and the Excite founders come to mind.

With all these new frontiers would be interesting to see where we are now over 3 hrs of well crafted PBS programming.

I think this is a killer ap for LIVE tv. Think about it, a superbowl game or a cricket match in india or a soccer match in europe...and you broadcast it over this technology, ( with ad breaks, so everyone is happy ). First- you wont have too many people wanting to save it and watch it at their leisure. Secondly, you would get a huge number of local viewers packed close together. It wont be like you have 10 viewers in brazil and another 5 in China, but more like 500 in the same neighbourhood, very effective for P2P purposes.

Neocast seems to be missing one point here. Most modern DSL is asynchronous and has much lower upload bandwidth than download. That means that a lot of "broadband" access won't have 700kbps of upload bandwidth to be able to forward the stream to the next person with. What will happen is that other viewers with more upload bandwidth will have to take up the slack and/or the content provider will have to from the source. This will still save bandwidth for the content provider, but I don't think it will plateau at 3-4 times the bandwidth of a single feed.

Also, it will still be possible to save the stream off to HDD since the entire stream will be available at some point or another. I'd say it's only a matter of time before someone cracks the protocol and writes a program that has the capability of dumping the stream to HDD as it is being viewed.

Correction: multicast routers are way below any application layer, they don't do caching and retransmitting video - that would be e.g. caching local/regional servers on the application layer.

Admins turn multicast off, because with multicast it's very to botch up your site's routing or being flooded in every subnet.

Admins turn multicast off because they always have.

What about firewalls and ports? Does this thing 'just work' on Mum and Dad's PC?

Very good article but what is the difference between Neokast and Joost?

Check out http://www.vividas.com/showcase.php This is almost 'instant on' It loads some very small player software and then begins to stream full screen! Quite amazing on regular 1,500 broadband.
Australian technology apparently.

Although Kahn and Cerf produced TCP/IP, the underlying network protocols of the internet, they certainly didn't have telivision or multicast in mind.

IP Multicast was invented much later, by Steve Deering, and neither Kahn nor Cerf have played much of a historical role in its development.

I don't see anything here that hasn't been around for years. When RIAA first took aim at internet radio with their first round of damaging rates, peercasting audio looked prepared to take off. The rates weren't quite bad enough to destroy streamed audio, so it didn't.

http://en.wikipedia.org/wiki/Peercast

The one missing ingredient: add a server that takes video calls from UMTS/3G phones and sets up a Neokast for whatever is send - and for a few hundred dollars we can all have "portable news links" - and the next revolution in citizen journalism.

If this is P2P + Multicast combined (which I think it is) then it is just more efficient P2P streaming. What is the probability of catching a multicast stream at the exact point I want to? Don't have a clue on the statistics - will be good to have.

But here's an interesting question/issue: For a potential viewer - how is this experience better than, say, YouTube, or iTunes, or any other method of receiving video data? Yes, I understand the live stream nature, and from a technical point of view it's a worthy, clever solution. The key to Neokast's eventually success (or lack), tho, is going to be how it's applied to consumers. Any thoughts on that, wise Mr. Cringley, sir?

@Tom. If Neokast requires vastly less bandwidth over important bottlenecks in the net, then the end user benefit is obviously a more reliable viewing experience and a higher resolution(if the broadcaster chooses to use the extra bandwidth possible)

It sounds like they are trying to do what another company called NetworkFoundation Technologies,llc (www.NFT-TV.com) is already doing. Cringley wrote about them sometime ago.

It sounds like they are trying to do what another company called Network Foundation Technologies,llc (www.NFT-TV.com) is already doing. Cringley wrote about them sometime ago

That's news?

Been there, done that. Lots of germany's soccer fans use p2p streaming tools to watch live streams from china.

http://www.pplive.com/

Download client, look in the list which of those curious channels has live German soccer matches, enjoy.

Easy as pie, and already an old hat.

Cheers!

How do you intend to have routers maintain these many S,G states? Have you ever seen the impact of several thousand on a router? Better yet, what happens when a link flaps?

People have been working on multicast for years and trying to get it to similar to cable quality. The problems still exist: Convergence, How to handle errors in transmission and Scale.

This is not innovative or unique and without the support of service providers, it wont happen. Figure out a better inter-domain multicast protocol (i.e. a faster way to converge), figure out how to deploy anycast'd caching servers rather than having to join back to individual unique source (since S,G's wont scale), and its plausible.

Just because something is streamed to you doesn't mean you can't save a copy for later use. If the packets land in software on my computer, I can always "crack open" the program that captures the bits and make it save a copy.

One of the important points of multicast is that there if there's DefectiveRightsManure, the key has to be the same for all players otherwise you loose the efficiency of duplicate packets. Once the key is exposed in software, it can be compromised almost immediately.

http://viidoo.com/en/index.php

Uses the TVUPlayer which is a standalone player which has built in peer to peer tv station streams. Works pretty well.

I am sorry to say it, but I've had to stop reading Cringely. The last half dozen columns or so have all been about the same thing: TV shows. I just can't get hyped about new and innovative ways for people to get their Barney Fife fix. There's got to be more going on in the world of high tech than that.

Neokast will support click and play On-Demand in addition to its live streaming capabilities. It is intuitively just live streaming from a file, with fast-forward (not for live streams obviously), rewind, pause, whatever. This means effectively no buffering time before viewing, as Bob indicated (2 seconds). This stream will run at incredibly high quality, and it will run indefinitely 24/7 with stopping if there's content to be streamed.

NFT is at the booth next to ours at Video on the Net in San Jose. I will give guest passes to the first ten people who respond today (I have to have you registered by the end of the work day today) to questions@neokast.com if they want to come to VON and compare quality, bandwidth reduction, connection time, buffer time, latency, and anything else between NFT and Neokast. PPLive illegally and intentionally distributes content without permission. They offer quality that is not even pixelly, it is blocky. Joost is not the same thing at all.

If you sign up to be a Neokast beta tester at www.neokast.com I will offer one month of free Neokast Pro services which will be released later this summer to this first fifty people who respond to our activation email with a contstructive comment about a technology like Neokast's.

Adam
President, Neokast

I guess not because you sure read this one, Tony.

Joost has massive partnerships and deep pockets. NFT showcased last year at the VON Boston in September and was very impressive (as well they have customers all over the open internet).
So if you are doing something better than Joost (who did I mention has deep pockets), or NFT (whom has proven its technology) then feel free to explain.

With all due respect and I am going to try and be constructive, but you have given us no “real” explaination of your technology, just “hype” statements about what you are going to do. You do not even seem to have a website, just a blog. So forgive me if I seem cynical but so far NeoCast sounds like another Zatto, Rawflow, Abacast, Octoshape, Chaincast (remember them)or BlueFalcon. These were the many predecessors that made great claims (and great press release, and some even raised money) but could never do - or even explain how they were going to do what they claimed- for that matter.

Does that win me the free service from Neocast that will be coming out really, really soon?

The problems are legal, not technical.

Go back and READ your ISP's Terms Of Service. Especially for broadband, there is one particular clause that is nearly ALWAYS in there - no servers. Peer-to-peer is nothing more than a client-server combination, and what the ISPs don't want are liability, legal complaints, or significant amounts of uplink traffic. How do you differentiate legitimate uplink traffic from serving pirated materials? "Good" subscribers just uplink requests for larger downlinks of information, so it's easier to forbid servers.

I want to know, does this technology preserves the line 23 Closed Captioning signal? Cisco IPTV and Apple ITMS throw away those bits! This is an embarassing shame, not just for the Deaf and Hard of Hearing, but also for families, like the Cringelys, who have young children at home.

I would certainly hope this works.

Because I can see this as being the savior for rock bands and other acts concerned about file downloading and the demise of the CD.

The future of rock music is live concerts delivered over the Web by subscription. Along with some personal touring, this concept eliminates the high cost of touring and allows the band to maintain a high level of contact with fans between albums.

This technology would make it possible without having to spend a fortune on bandwidth costs.

As for the blocking of "servers" by ISPs, this is a minor issue. It may prevent everybody from being a broadcast TV station for free, but it obviously is a revenue opportunity for ISPs who SELL the bandwidth and technology to allow people to broadcast.

Obviously ISPs providing the bandwidth should expect to share in the revenues generated from using that bandwidth, "common carrier" status or not. By allowing them to charge more for uploads or broadcasts, this can be done, provided they allow a certain degree of broadcasting from people who *aren't* getting revenue from their broadcasting.

That way, everybody wins. Nonprofit people and ordinary users can do a little broadcasting for free, profit organizations can do a lot, and both the profit organization and the ISP make money. The rest is up to competition (assuming the Feds don't reinstate the ATT monopoly.)

Since this is a .NET app, it should work under Mono; it shouldn't be limited to just Windows platforms. Unless the author makes use of GDI+ or DirectX API's to display the video.

The "Cringely Does Neokast" video file has the filename extension .FLV. Windows Media Player 11 cannot play it, and I doubt very much that Real Player 8.x will play it. So, just what specialized video player do we "need" to play the video file of the interview??? In your opinion, how many video players should an ordinary user have installed on their PC to validate the egos of the video developers??

The "Cringely Does Neokast" interview video file has the filename extension .FLV. Windows Media Player 11 does not play it, neither --I suppose-- would Real Player 8. So, which specialized video player do we need to download and install just to view this one file probably no more than one time??

Access Curmudgeon: there is currently no easy way to preserve a line 23 CC signal when compressing to an internet-friendly digital format, sorry. You could decode it to text and convert it to a subtitle format for inclusion in an .mp4 or proprietary-format file, but there's not enough interest yet for big players like Apple to bother supporting it across the board. "Internet TV" whether downloaded or broadcast is still in its infancy and you can't expect niche features like CC-subtitles, useful to less than 1%, to get much support yet.

There is a TV streaming service here in Switzerland called zattoo.com. IMHO they use exactly the same P2P technology as you are describing for Neokast. They are online since several months already...

Why are you wasting time writing this useless stuff? What happened to NerdTV. I urge you to focus on that and bring it to fruition. Thanks.

A company called Abacast has been selling a very similar technology in practice for at least five years. They even have customers already. Sure, it's not as sexy as a new startup, but it's market-tested and probably Nerd-TV ready.

You could easily fill 24 hours a day, just broadcast "Plane Crazy" on a loop.

:)

Ryan

Good god, Robert.
When will you finally realize that no-one gives a fsck about TV over the internet?
We already have a perfectly good delivery mechanism for TV, one that allows me to record shows and that works with my existing hardware. Why would I give that up for something that I can't record and that provides an inferior user experience.

We already know what people want to use the internet for. They want to be able to choose what they watch and when they watch it --- podcasts and YouTube, with podcast the preferred choice because they don't tie you to a PC.
You sound like those self-deluded morons who every so often pop up to tell us that their commercials will be so compelling that people will want to watch them rather than skip over them. Just because you repeat it a thousand times doesn't make it true.

I think some of you miss the point. The point is cheap broadcasting. Of course, the content needs to be interesting, but thinking of Youtube, the public demand is not that demanding. So, even if you, as a consumer, don't care much for IPTV, many content suppliers will. And if you want to see their broadcast, You have to accept doing it on your PC. Then -- why Neokast? As long as nobody else has made a success out of the idea, the way is free for anyone who might offer working technology and a working business concept.

I'm with Troy and Maynard. Enough about TV over the 'net already. But what should we expect here, really... double-check the URL: pbs.org

I'm with Troy and Maynard. Enough about TV over the 'net already. But what should we expect here, really... double-check the URL: pbs.org

(sorry for the double-post... server gave me a "whoops no such page" error the first time.)

I've been watching live streaming with ESM for years! Go CMU! Albeit there's not much content on there...yet. http://esm.cs.cmu.edu/

Hey! I thought of that idea back in 2001!

It was my engineering honours project.


http://www.trevbus.org

Well done to whoever can get it to work, and get it taken up by the masses.

I would like to suggest the main reason ISP did not turn on multicasting was because it would decrease the amount of bandwidth their users needed, not because of stress on the routers.

Cisco routers and switches can easily keep track of the mulitcast groups (streams), it was the ISPs who don't want you to have 100 people watching a 700Kbps stream.

"Neokast is scrupulously polite."

Detail: This particular Neokast peer is polite, but the real questions is whether the protocol is vulnerable to rude peers (think STMP). A few rouge peers that hit the primary server when they don't need to, or don't announce themselves on the local network, or worse do announce themselves but serve something other than the original stream (video spam!) could bring the whole thing to it knees.

kz

kz-

I don't know anything about Neokast, but if I were designing the system, I would probably make client authentication a requirement and keep that handshake closed to ensure that only legit clients made the connection. Additionally, SPAM shouldn't be an issue in a torrent-like multipoint peer-to-peer stream. Anyone beyond the first client to connect to the server will receive their stream from multiple sources to ensure that outbound bandwidth requirements are minimal per P2P client. Even if someone gets a rogue client on the network, all they could do really is disrupt the stream with bad data, but they couldn't provide their own show, as their data would just be interleaved with data from the real stream.

Of course, I'm making some assumptions here, but I'd say that this is a real likelihood.

This is certainly not new stuff. There are many other P2P technologies out there already. I recently reported my experience watching the recent Superbowl 2007 over P2P (http://www.broadcasting20.org/2007/02/05/p2p-superbowl-2007/).

I think they should have called the company MagicKast rather than NeoKast as it is physically impossible to push 700Kbs to your peers if your upstream is only 256Kbs (like mine). Where is the missing 446kbs (per-peer) coming from? And this if you only have one "child" in your peer-to-peer mesh. What if you have 2 children? With only one child you would end up with a nice long chain which as soon as a node in the middle leaves the network would leave all the dependant nodes with no stream... I am very puzzled...

What do you mean: Where is the missing 446kbps coming from?

It's peer to peer. Clearly, the initial seed stream would need to come from somwhere that has more bandwidth than a typical cable or DSL upstream. However, if you need 700kbps, and there are 10 other people already in the stream, you only need 70kbps from each one, which is well within the limits of most broadband upstream limits.

Wow, The network IS the computer. You heard it here first....

I'm wondering if there is sthing new in Neokast solution compared with already "old" P2P tv stations like tvant, pplive, etc.


_

I don't know how happy your ISP will be when you start downloading 50 Gigs / month to get average 5 hours of TV per day. That traffic will get blocked quick no doubt.

Why would finding streamed video content in a million-channel world be any harder than finding content in a billion-website world? No confidence in video.Google.com??

Cringley, you have to look at www.reeltime.com
It uses a P2P network just like the Neocast one you refer to, yet it has begun to organize films and TV programming on its site. The P2P network it uses is a Seattle company called GridNetwork's Powergrid player, and it seems to work just as you are describing the Neocast network..
Check out the films and the cheap subscription prices. They will have PPV on the site shortly from Lionsgate and the CHUM TV Network in Canada.

www.reeltime.com is up and running and is even better than neocast, they are loading Lionsgate films as we speak. Test one of their free to watch DVD video trailors, Reeltime is carving out a big piece of this market...

I think you missed the point for once, Cringe. Neokast is an idea, and not an original one, or a useful one, at that. Channels of video are so yesterday. Video on demand is the future. Watch what you want, when you want it. Point click and watch is the way to go. And it is available now, with all the p2p features you are so enamored of. The two comments right above point to a great site for that. You covered them way back in March of 2006!

Michael, Your are partly correct. Neokast is an Idea, not original and they have no product or customers.
As far as Channels of linear content....well thats where you are lost. Take for example radio, when people drive all the time in there car where they have mp3 players, cd's and dvd players, etc. And over 80% of the time they still listen to the radio. That is because people like to be programmed to, and people like to feel connected to the world (even when that channel is not live like the JACKFM format for example). And there are similar stats on DVR not killing broadcast channels on cable.
Now before your start typing a response I am not saying there is no use for on demand- clearly there is, but it is a much smaller market then the market for linear channels.

(oh yea did I forget the main reason that people are generally lazy!)
;)

To Rob: Re: ABACAST:
It is very well known in the industry and admitted in the press by their CEO in the press ("You can only make false promises for so long before it starts to catch up," King said - http://www.columbian.com/business/businessNews/02042007news100525.cfm) that although Abacast has said they have this technology they simply use it as a bait and switch to just sale traditional Unicast web broadcasting. I have never heard of anyone using it to stream live video, and it is my understanding that what it does for audio is limited because their technology is fundamentally flawed at the architectural level

Finding interesting content in a million channel world couldn't be much worse than finding interesting content in the roughly 100 channels I have on my cable today.

Mostly there are 100 uninteresting programs.

There was an interesting company that developed an AI application for parsing video. The idea was that keywords and scene shifts would automatically be used to index video streams. You could search to find the segments that you were interested in, instead of being forced to watch the whole thing.

They had sold it to CNN and others back around 2000. Great for news material! Great for 'cutting to the bottom line'.

William

I WANT A COMPUTOR VIDIO TECHNOLOGY WHICH ALLOWS ME TO SEE TEH FUCHING FERRET WHO IS STEELING THE LARGE WHEEL OF CHEESE FROM MY HOUSE AND ROILING IT DOWN THE STREET

Just had to note that Linux will run .Net applications (some at least). See http://www.mono-project.com/Main_Page.