Podcast: Play in new window | Download

There’s a dispute going on right now between Comcast and Level3 Communications concerning the peering agreement between those two companies. Comcast says the dispute has nothing to do with the fact that Level3 just got the Netflix video streaming contract while most observers think that’s all it has to do with.

I think so, too.

Peering is at heart nothing but restraint of trade. Peering came about when various Internet backbone providers noticed they were all connected to the same big data centers and points of interconnection, normally inside telco central offices. Simply pulling an Ethernet cable from one rack to another could interconnect millions of users from two different backbone providers, saving time, distance, router hops and total bits in the process. Peering agreements typically involve no exchange of money since they are intended to be between peers — very similar companies of roughly comparable size that would be sharing equal numbers of bits back and forth. Peering agreements were for big companies, especially backbone providers interconnecting with the fundamental idea that they’d be giving as many bits and they got and therefore no direct compensation would be required. It also kept smaller companies out of the backbone business because they were made to pay, and dearly.

Level3 is mainly a backbone company that is lately delivering a lot of streaming video, too. Comcast points to the disparity between the number of Netflix video bits served (a lot) to the number received (almost none for Netflix other than some Quality of Service data and of course the movie orders). That’s not the deal, says Comcast, which wants Level3 to pay the difference in cash.

On the other hand, Comcast for the most part isn’t an Internet backbone provider. They have some backbone assets, sure, but mainly they are America’s largest broadband ISP. So while Comcast can fault Level3 for taking advantage of their peering agreement terms, Level3 could as easily drop peering with Comcast altogether, still getting to Comcast viewers through other peers, though with the addition of some latency from the extra hops required.

Note that Netflix formerly did its streaming through Akamai’s Content Distribution Network (CDN) which shares revenue with participating ISPs.  Level3 probably got the Netflix gig by beating Akamai on price and they beat Akamai on price because they are relying on that darned peering agreement to make it possible.

As an ISP, Comcast could afford to drop one backbone, but not all of them, so Level3 has some power here — more than many commentators have noticed.

There is a lot of posturing here, so let’s try to figure out the real issue, which I think is Google.

Google has long wanted to drop a rack or a container or at least its own fiber connection two hops from every broadband user in Ameica and eventually the world. They’d like to do that through peering agreements like Level3 and certainly have as much of an argument as Level3 has for doing so, given Google’s own fiber assets, which are certainly more than Comcast’s. But Google will pay for access if it must, because global domination is worth the price.  The search giant is willing to pay if it must for guaranteed access.

Comcast knows this. As America’s largest broadband ISP, Comcast stands to gain more than any other company from allowing Google to run fiber into every head-end data center the company has. But Google won’t pay if they don’t have to. So to make sure Google pays, Comcast has to make sure Level3 pays.

That’s all it is. Both sides are distorting the peering agreement like crazy to make their points, which aren’t about equity, net neutrality, user rights, legal theory, who is actually paying for the bandwidth (customers), or anything else — just Google’s money.

Podcast: Play in new window | Download

What happened to Neokast?  It’s a mystery to me.  But I suspect the answer will surprise us all soon enough.

Neokast, as readers of my old PBS column will recall, was a peer-to-peer live video streaming application developed by graduate students from Northwestern University near Chicago.  That’s me talking about it there on the left, back in 2007.

I loved the company instantly. It was out of the Silicon Valley limelight, away from the technical mainstream for such software (Neokast was a .NET application and therefore pretty much Windows-only), but most important of all, it seemed to actually work.  The potential was extremely compelling.  Here’s how I described it back then:

“…the more people who watch your Neokast the more efficiently will your server bandwidth be utilized. According to Birrer, under normal circumstances the server bandwidth should plateau at 3-4 times that of a single stream NO MATTER HOW MANY VIEWERS ARE BEING SERVED. With a per-stream bandwidth of 700 kilobits per second, this means that Neokast would never require more than a continuous three megabits per second of server bandwidth per video channel. Let’s put that in a real-world context. Three megabits per second is almost precisely 1000 gigabytes per month, which is half the allotted monthly throughput for a $6.99-per-month web site at 1&1. So if Neokast’s claim is valid, it would be possible to broadcast American Idol or the Super Bowl or friggin’ CNN worldwide for $7 per month.”

I called it “The $7 Television Network.”

Response to that column was electric, as was the reaction to Neokast, itself, when the beta software was shown shortly thereafter at a trade show in California.

Neokast was on a roll.  Now all they had to do was deliver.

But apparently they didn’t because now Neokast is gone.  Their web site is dark.  The entire technical team, as far as I can tell, left last September to start a new company in an new product space – content management.  The company’s sole patent application even lost its legal representation in January when a Chicago-based law firm withdrew.  Only NeoKast CEO Adam Johnson remains with the company and that’s only according to his Facebook page, where he appears with a remarkable variety of very attractive young women.

It’s a familiar story, right?  The idea was good but the code wasn’t.  OR the code was good but they ran out of money.  OR the code was good and they had enough money but the founders had a falling-out.  OR any other mundane reason that you might care to come up with.  Neokast is gone, so what?

I’ll tell you so what.  My 32 years in this industry tell me that none of those possibilities is true and that some aspect of Neokast is alive and well, though probably under a different name.

I put these guys on the map.  I wrote about them in a way that gained them a huge amount of attention at a time when they were getting no attention from anyone.  All that means, really, is that their mothers would teach them to be nice to me.  And sure enough, on January 28th, when Adam Johnson and I happened to share a birthday, though almost 30 years apart, we wished each other well within hours on Facebook.

But Adam DIDN’T tell me then that his company was effectively dead.  It took a reader to point that out to me a couple weeks later.  And when I went down that same Facebook path I’d used to wish Adam a happy birthday — this time to ask what happened to Neokast — I got no reply at all.  So I tried again, more forcefully.  Still no reply.

So I tried a couple of the ex-Neokast technical guys at their new startup.  No answer.

No answer?  Don’t these people want to promote their new technology?  They know what I can do for them; don’t they want me to do it again?

No answer.

This doesn’t happen, not to me.

So what’s the deal?  I don’t know.  But I have a theory.

I think Neokast was bought for a lot of stock or money by some well-known company. The way the technical team was handled in this transaction it looks like the acquirer wanted the code, not the coders, which suggests a company with confidence, even arrogance, and technical depth.

The reason I can’t get anyone to respond is simple under this scenario: they have to all be under a particularly onerous non-disclosure agreement that will take back the money if they say anything – ANYTHING – about the deal.  They aren’t prohibited from just discussing it, they are prohibited even from ACKNOWLEDGING it, hence the total silence.

At one point last year I was told that Microsoft had made an offer for Neokast but was rebuffed.  So maybe Microsoft came back again, this time with the BIG checkbook.  I think this is most likely.  But there are other possibilities.  Apple could have acquired Neokast to kill it.  IBM could have acquired it to become a player in the streaming video business.  Or Sun.  Or Cisco. Or some other company, up to even a Comcast, though I don’t see the cable company as being so techie as to rebuff the coders.

If you’ve been paying attention to entertainment news you may have read lately that there is a lot of shuffling for position going on between cable companies, telephone companies, cable and broadcast TV networks, and various startups for dominance of  live or on-demand TV channels over the web.  But all this talk so far seems to be based on using content distribution networks, not peer-to-peer.  Even Joost, the p2p video site from the founders of Skype, has publically given-up on using peer-to-peer distribution, leaving only Grid Networks, as far as I can tell, ostensibly in the live p2p space, if just barely so.

Let’s guess for a moment the acquirer IS Microsoft.  Because there has been no public announcement of such an acquisition the buyer has to be a big company like Redmond, where the size of the deal wouldn’t be considered “material” to their business, so they could avoid being required by the SEC to even issue a press release.  Of course it could just as easily be the other suspects I named.

But there are many reasons to believe the buyer is Microsoft.  They took a run at the company before, remember.  They are perfectly equipped to handle the technical job on their own, provided they keep the number of hands to a minimum.  They could still screw it up.

There’s no indication, by the way, that Microsoft has done this.  Certainly none of my friends who know Microsoft have heard anything.  But that could just mean they changed the name and Neokast is now MicroKast or some such thing.

Now here comes Windows 7 – a perfect place to stash a Neokast p2p client.  On the other hand, Microsoft could put Neokast code in its regular monthly update and get it running on 20 million .NET nodes overnight.  That’s what I would do. What’s funny is if a startup did that there would be an uproar about security, but if Microsoft deploys Neokast overnight it will be seen mainly as a clever move.

Microsoft is desperate to have something new to control and media distribution is their target.  They want to control movies and television the same way they have long controlled software.  But right now Apple and Google are both doing better than Microsoft is in this space.  Ballmer will do anything to beat Apple and Google and the only way to do that – the ONLY way to do that – is by introducing some new game-changing technology like massive, really cheap, delivery of LIVE video.  Bring the TiVO video experience to the World Wide Web without requiring a cable box OR an antenna.  Well Neokast takes a good shot at doing just that.  And bringing the horsepower of 20 million servers to the task would make it even easier. 

Maybe the acquirer isn’t Microsoft.  But I’ll tell you right now that some big company somewhere has snapped-up Neokast, is continuing to develop the software and intends to introduce it soon with a big splash.  I just wish I knew who it was.