Remember how that freed-up spectrum was up for auction and Google made loud noises about bidding. I even predicted that they would bid, because that’s what I was hearing from inside the Googleplex.  Google wanted to set up a national wireless network to rival anything from Verizon or AT&T. Only Google didn’t follow-through on its threat to bid and the frequencies were cherry-picked, instead, primarily by the big wireless incumbent carriers who have for the most part done little with them.

They bought the spectrum primarily to keep it out of play, to keep a viable competitor from emerging.

That decision not to bid back in 2008 seemed very short-sighted of Google. But now I hear from people who were inside the FCC at the time that Google was privately told by the Bush Administration not to bid.

What if Google had defied this government nudge? I guess the threat was they’d have it taken from them anyway through some regulatory action or legal challenge. But had Google succeeded, we wouldn’t be seeing bandwidth caps being imposed today on wireless data plans. And wireless data would be cheaper everywhere.

Our tax dollars at work….

Dear Steve,

I don’t think either you or President Obama will like what I write here so you may not choose to use it.

President Obama has a message problem, true. He’s simply out of his depth.

Here’s a guy who has always been the smartest in the room, noted for his Teflon smoothness and turn of phrase. But in order to turn the right phrase, no matter how smart you are, still requires an understanding of context — an understanding he simply doesn’t have. And why should he? Obama was always a political operative. Michelle has more real world experience.

He’s not dumb, he’s not shallow, but he is inexperienced.

Shared sacrifice (and calling for it) works when you have no easier choices, when the alternative isn’t just a different policy but military defeat and potential annihilation. That’s why Churchill could make so well those very calls that Obama is frankly attempting to emulate now. Only this isn’t the Blitz — it’s the Great Recession — and the GOP isn’t Hitler, just a bunch of pols. So this was a misstep on Obama’s part.

But you are wrong, too. You know a lot of entrepreneurs, but I’m not sure you have ever been an entrepreneur. I have. I started and helped to start several companies that failed completely. I also helped to start three companies that have a combined market cap this afternoon of more than $500 billion.

One question I am frequently asked because of my background is if this is a good time to be an entrepreneur — a good time to start a company? Understand people have been asking me this question for 30 years — a period of time that encompasses some major booms as well as two of our deepest recessions. And the funny thing is that my answer to this question never changes: it is always a great time to be an entrepreneur and start a business. And with the passage of time I tend to think it has only gotten better, even today.

The sort of entrepreneurism I know, which is technology entrepreneurism, is acyclical. It doesn’t matter what the economy is doing, what interest rates are, even the mood of the venture capitalists. People will still start new technology companies to develop their great ideas. Understand that they do so with a 95 percent chance of failure, again no matter how the economy is doing. Against an obstacle that huge, factors like government policies and political talking points are inconsequential.

Entrepreneurs start companies compulsively. It would take medication to force them to stop, thank God.

So what we have here with President Obama and Congress is at best positioning and at worst posturing and it will have no impact whatsoever on entrepreneurial energy in this country.

Entrepreneurs have more important things to worry about.

Not so sick after all?

This may seem an odd topic, but stick with me. Yesterday Disney’s ABC television network said it was licensing two canceled daytime TV soap operas to a production company that would be moving the shows to the Internet. I seem to be the only one who thinks this is a brilliant move. In fact it might be the Internet’s next killer app.

All My Children and One Life to Live as killer apps? Yes.

A killer app, remember, is the Silicon Valley term for an application that all by itself justifies to certain users the acquisition of hardware needed to run that app. People will go down to the store and buy hardware just to be able to use that application, whatever it is. VisiCalc was the killer app for the Apple ][, Lotus 1-2-3 was the killer app for the IBM PC, Halo was the xBox killer app, and Bonanza was the killer app for U.S. color TV.

Every new platform needs a killer app to get beyond the early adopters and reach a broader audience. Some experts are arguing that Netflix is the killer app for Internet TV, which might be right. But I think it might just as easily be All My Children.

ABC had nothing to lose in this deal. The two soaps in question were already canceled, winding down their casts and story lines. I’m sure they were acquired for a dollar each or less. If moving online saved ABC some costs from closing-down the series they may have actually paid Prospect Park to take the shows off its hands. So these iconic brands were cheap to acquire.

But they are expensive to produce, right?

Not so fast.

Nobody outside the shows and their networks knows for sure but a $50 million figure is often thrown out as the annual production cost of a major soap, so let’s work with that.

Fifty million dollars is $192,000 per episode or $4,370 per finished minute based on 44 minute shows. That’s a lot of money but a lot less than primetime TV budgets. It’s also the absolute most any soap has ever cost with most costing less. Certainly there are some savings to be found in there. Let’s claim a 20 percent labor savings from moving to the Internet, bringing per minute costs down to $3,496.

Actually, there are plenty of additional savings. Some savings will come from lower labor costs as actors accept smaller paychecks as an alternative to retirement or unemployment. But an even greater savings will come from any Internet soap’s ability to offer online every episode ever broadcast — the long tail — at an effective production cost of $0 per hour.

If a third of Internet viewers are watching old episodes that drops the effective cost of new episodes by a third, so we are down to $2,342 per finished minute.

Don’t forget potential subsidies from hardware companies. As a killer app for Internet-connected TVs, for example, All My Children might get some cash from TV manufacturers. Bonanza got money from RCA that way and many popular shows were moved to HD production with financial support from HDTV makers. Why not the same for Erica Kane?

Our All My Children budget is now down to $26.8 million per year, so let’s figure $4 million of that might come from Samsung or Panasonic or maybe even Google TV if any of those platforms can be somehow uniquely linked to the shows, possibly through additional or interactive content.

Heck, what if All My Children could be accessed solely through its Facebook page? How much would Mark Zuckerberg pay for that?

I don’t know what Zuckerberg would pay, but I do have one number to work with — the rumored production budgets at YouTube’s upcoming professional channels. According to Variety, YouTube will shortly bring some professional channels to its service with budgets of $1000-$3000 per finished minute.

Our straw man budget for All My Children, which now stands at $22.8 million per year, just happens to work out to $2000 per minute — right in the sweet spot of those rumored YouTube numbers.

I am not saying that All My Children and One Life to Live are headed to YouTube as the basis of a Soap Channel, but I am saying that they’d be profitable both for their producers and for YouTube if they were headed there.

Each show has about 2.5 million daily viewers — each a potential buyer of an Internet-connected TV. That’s $2.5 billion worth of TVs and well worth a $4 million production subsidy.

If YouTube or any of its competitive services could reliably get 2.5 million viewers per original episode they’d see that as well worth the money, too.

This is long form video with commercial breaks going to a dedicated audience which can now be global (that last part could be huge). Remember 2.5 million viewers of a 44-minute soap opera is the equivalent of 36 million typical three-minute YouTube video views. As professional content with a 40 year heritage that’s an easy sell to advertisers — a no-brainer for P&G.

So contrary to all the skepticism, moving soaps to the net could easily become a goldmine — one with a lifespan far longer than that of VisiCalc.

Democrasoft’s Collaborize product began as a way for communities to discuss issues online with the idea that the core groups would be cities or local governments. But the Santa Rosa, CA-based company found some of its earliest adopters were teachers — a group the company had never even considered. More than 6600 teachers are presently using Collaborize, which is still a very small percentage of American classrooms. But the adoption growth rate is a very viral 30+ percent per month, according to Democrasoft CEO Richard Lang, so this application is going to be significant.

A lot has changed about Collaborize since I last wrote about it. The business model for education is now free, for example, and there is a custom version, Collaborize Classroom, just for schools — the company’s largest present market.

The idea behind Collaborize is simple. It is a structured conversation. Any participant can pose a question for discussion, eliciting responses from the group or taking a poll. You can look at it as a quiz or a test but grading doesn’t have to be a part of what’s essentially an online Socratic dialog.

Yes, you can do this with a wiki, Mr. Smartypants, but the significant point here is that people generally don’t do this with wikis. Collaborize is more structured than a wiki and requires little customization.

One of the most important aspects of this tool, according to the teachers who like it, is that it has a social leveling function that brings students into the online conversation who might say little or nothing in the classroom. This is good.

What’s changed most recently with the product, though, is the addition of some social networking functions and especially the development of a library of discussions available to anyone.

Here’s the deal with the library. Some teachers are better than others at designing discussions. These can include video and audio clips to spur discussion and each becomes a little lesson in its own right, providing information, eliciting responses, and even measuring comprehension all at the same time. That’s powerful. But if you are a teacher who is intimidated by the whole question design process, why not use some other teacher’s discussions, either in their entirety or as the basis for your own derivative work?

The library already contains thousands of discussions in nearly every subject area.

Social networking provides both the door to this library and something even more important — peer review. Maybe a discussion could be improved or maybe it is just plain misguided: responses from other teachers will provide that context. If a discussion from the library gets a lot of positive responses from teachers whose opinions you respect that makes the discussion more valuable.

I think this is great, especially because we Cringelys are embarking now on our own experiment in home schooling, where I’ll be borrowing shamelessly from Collaborize Classroom (my only connection with the company, by the way).

The more tools I have on my belt the better I feel.

I am thinking of consolidating all my domains with one registrar. Where should I go and why?

Google is the youngest of these companies and has probably the smallest patent portfolio, most of which isn’t mobile or telecom related. This puts Google and Android at a legal disadvantage and explains the 45 patent infringement suits that one analyst says Google in presently facing in the mobile area alone.

Google would have preferred to win the auction, but with the consortium sitting on more than $100 billion in cash, the outcome came down to determination, not resources. Google stayed in it only long enough to make sure of the consortium’s intentions and to make the purchase more painful for them, if that mattered.

It certainly mattered to Google, because that $4.5 billion number will be at the heart of the inevitable anti-trust lawsuit Google will file almost immediately. Every good anti-trust lawyer in America just cancelled his or her July 4th holiday to prepare their pitch for Google, which will probably claim Restraint of Trade as well.

Given that the courts will shortly be involved, Google can probably operate unfettered for another 2-3 years, during which they’ll try to build their own mobile patent portfolio. Google may well be able to use the courts to slow the actual Nortel transaction, too, according to my lawyer friends.

So the “Android is dead” story here is way premature.

In the long run, remember, Google will probably be able to use its legal strategy to force the consortium to at least license some or all of the patents. They’ll get a royalty from Google, I suppose, and thus benefit from Android’s success, but then Google is unlikely to be completely deterred, either.

The story everyone seems to be missing here is who gets what in this consortium deal? Most journalists and bloggers seem to assume the winners will all share equally in the IP spoils. But I have people who know people and the word I am hearing it that’s not the way the consortium works at all.

Some consortium members get patents, some get royalties, and some just get freedom from having to pay royalties.

Notice Nokia isn’t in the consortium? The Finnish company is apparently covered by Microsoft, tying Nokia even more firmly to Windows Phone.

Here’s the consortium participation as I understand it. RIM and Ericsson together put up $1.1 billion with Ericsson getting a fully paid-up license to the portfolio while RIM, as a Canadian company like Nortel, gets a paid-up license plus possibly some carry forward operating losses from Nortel, which has plenty of such losses to spare. For RIM the deal might actually have a net zero cost after tax savings, which the Canadian business press hasn’t yet figured out.

Microsoft and Sony put up another $1 billion.

There is a reportedly a side deal for about $400 million with EMC that has the storage company walking with sole ownership of an unspecified subset of the Nortel patents.

Finally Apple put up $2 billion for outright ownership of Nortel’s Long Term Evolution (4G) patents as well as another package of patents supposedly intended to hobble Android.

At the end of the day this deal isn’t about royalties. It is about trying to kill Android.

Note — Here’s a pretty good account from Reuters of the Nortel patent auction. You’ll notice they don’t include the participation breakdown of the winning bid (who gets what) that so far appears no place but here.

I called you about this but you didn’t return my call. Your CNet TV site has been stealing my work.

Try a search on “CNet” and “NerdTV” and you’ll find many of my NerdTV my shows, complete with their Creative Commons attributions at the end prohibiting their use on commercial sites. CNet TV is a commercial site because it carries advertising.

Why does your professional media operation feel the need to rip off my work?

Here’s my Google search:

Here’s a sample CNet page. Notice the Blackberry ad.  There are other ads.

I don’t know how long this has been going on but stop it immediately.

These shows and NerdTV season 2 will appear in the fall on a site that actually asked permission to use them.

Take them down now.

Note — As of  Sunday evening July 3rd Cnet appears to have taken down the NerdTV videos.

Security gate from Startown Rd.

In a few days we’ll be leaving Carolina, possibly forever. Following the recent death of my father-in-law — our reason for coming here in the first place — Silicon Valley calls once more. But before leaving town I was determined to scope out that $1 billion Apple data center in Maiden, NC. So I drove over, took some pictures, and talked to folks at the convenience store down the road. My conclusions from this unscientific research is that the giant Apple facility is mainly empty. It’s a huge building filled more or less with nothing and why Apple built it that way frankly escapes me. Maybe it’s just a shot across the bow of Google and its $650 million data center in South Carolina.

The place is certainly locked down. There’s a security gate on Startown Road and not much else to see. Apple has built a tall earthen berm around the entire site and planted trees atop that. The only way I could find to see the data center from ground level was from the Baptist cemetery next door. All that’s visible is the top of a huge white building and a couple of big tanks that may hold cooling water or, more likely, diesel fuel for emergency power generators.

Not much to see

I parked across from the main gate for an hour during the middle of the day and one pickup truck entered the facility.

It’s not that Apple has so little to do at the new data center which, after all, is supposed to be the center of iCloud and iTunes activity, updating all those Macs, iPhones, and iPads while serving video and audio to more than 200 million devices in all. That job isn’t trivial, but how much square footage does it actually take to do?

For comparison purposes, I looked at IBM’s Special Events Web Service (the gang that used to do the Olympics). They have 2000 square feet in three different data centers. They have a couple of racks of cache servers that handle over 95 percent of the actual work. Behind the cache servers are three racks containing about 50 1U linux servers setup in a cluster. These manage all the transaction work and anything that leaks through the cache servers.

Maybe the Olympics isn’t a good comparison, but where IBM has 2000 square feet, Apple has one million square feet — 500 times as much.

According to the Internet Movie Database there are about 700,000 movies in existence, excluding porn. Most movies will fit (in DVD form) in 4.7 gigabytes. Do the math and you get 3290 terabytes, which is a big number but not that big. Most data centers serving media files would cache about 10 percent for optimum performance. That’s 329 terabytes. Knowing a good percentage of movies aren’t worth the film they were printed on, you can probably come up with a 50 terabyte caching design and be able to serve anything anyone would want to see. Fifty terabytes of cache servers can fit into a couple of racks.

When planning a data center each rack requires about six square feet of floor space. But for the sake of discussion let’s make that 10 square feet to allow for non-server areas in the building. One million square feet divided by 10 square feet per rack means 100,000 racks could be constructed in the Apple facility. That’s 7.2 million 1U servers unless the racks are built extra-high, in which case there could be more than 7.2million servers.

Remember that between the 3290 terabytes of disk storage, 329 terabytes of cache and all associated servers, load balancers, etc. we’re talking at most 20 racks to serve every movie ever made. Now increase that by a factor of 10 because I probably blew a calculation somewhere. Now increase that by another factor of 10 because Apple may want to serve not only all our movies but all our TV shows, too. That brings us to 2,000 racks — two percent of the capacity of Apple’s data center.

Are you beginning to get my drift here?

Now Google isn’t Apple. Google is continually indexing the whole darned Internet, runs the biggest e-mail service, and many other services, not to mention all those ads. You can see how Google would require lots of data storage and servers to handle it. By now Google must have over a million servers and a very elegant way to manage them. But even Google’s million servers would require only 13.9 percent of Apple’s data center capacity.

So what is Apple doing with such a big building? I can’t imagine a workload that would need even a tenth of that data center.

Maybe they are building for the future, you say.

That’s crazy. Remember Moore’s Law? As time passes all of those Apple racks will be filled with new computers that are faster and have more memory and with storage systems that hold more, too. The square footage requirements are, in fact, likely to stay about the same for the foreseeable future, absent some quantum expansion of Apple’s services.

For that matter, what servers is Apple using, anyway? Certainly not the now discontinued xServes. That alone may be the reason why they’ve made the facility so difficult to see, not wanting to boost any competitor by admitting Apple is a customer.

So here’s my guess: I think it’s a joke. The building is a near-empty facility built primarily to intimidate Apple competitors. And so far it seems to be working.

Death Star

I’ve been thinking about Apple’s App Store and the industry paradigm shift it represents. Apple loves to change the game like this, simultaneously unseating previously entrenched adversaries while building for itself a defensible system for the future. The trick to making it work is to not appear to be too greedy and I think Apple is accomplishing that. They are greedy, of course, but as Fernando used to say, “It is better to look good than to feel (or be?) good.”

Apple’s original App store was for the iPhone — a portable and for the most part cloud based method of distributing and updating iPhone apps. This was followed by Apple’s App Store for OS X, which did much the same for Macs. Both are being extended fully into the cloud next month with the release of OS X 10.7. For users the App Store lowers the cost of applications, keeps them updated and synced, and allows their deployment across several computers. For Apple, the App Store destroys shrink wrapped software, eliminates product serial numbers, vanquishes piracy, and punishes competitors like Adobe.

Software goes from being a box of bits to a cloud of electrons. Remember Larry Ellison railing against the box of bits metaphor in my show Nerds 2.01: A Brief History of the Internet? That was back in 1998. None of us, even Larry, knew it would take 13 years for that vision to be realized.

With the App Store prices are lower because costs are lower, but also because Apple wants prices lower to gain market share for both its devices and the associated ecosystem. That’s an important but little recognized part of this paradigm shift. The old question used to be whether Apple was a hardware company that sold software or a software company that packaged its products in hardware. The new reality is that Apple is an ecosystem in which hardware and software are important but then so is the cloud that lies behind both.

At the same time that the App Store allows you to run one $299 copy of the new Final Cut X on all your computers, it becomes nearly impossible to pirate that software without first hacking Apple’s data center in North Carolina. This is huge and its effects will be profound, keeping legit customers honest at little cost while pushing pirates toward other solutions, especially Open Source.

But what about Adobe or Microsoft or Symantec? They can sell their software through Apple’s store, accepting lower prices and sharing 30 percent of the money with Apple. Or they can stick with serial numbers and piracy. Or they can roll their own app stores, but in doing so forgo the power of the Apple ID or risk infringing Apple IP by somehow reverse engineering it.

It’s a tour du force that will have painful consequences for competitive products like Adobe’s Creative Suite. Apple to Adobe: we win, you lose.

This sin shall not go unpunished.

Among his milestones IBM’s VP of Innovation completely forgets to mention the company having helped automate the Third Reich.

And while IBM was celebrating other noteworthy achievements, a reader pointed out to me what he thought was an IBM data breach:

“My wife and I are Health Net customers. A month or so ago we received a letter from Health Net saying that their contractor, IBM, had been hacked and that our medical records including SS# had been stolen… You can imagine how I feel about it. I’m in favor of the bin Laden treatment for the hackers and serious bitch slapping for everyone else concerned, from the pointy haired managers to the OS pukes who have refused to create secure systems despite knowing how to do it. The people who have resisted IPv6, which provides authentication, over the last decade are another good target for serious bitch slapping. Someone said that the primary reason the computer industry advances is ridicule of second rate technology. Ridicule of insecure systems and networks is desperately needed.”

To be fair to Big Blue, it appears their system wasn’t hacked in the manner we’ve been discussing lately and IPv6 had nothing to do with it. Rather, in March IBM discovered nine disk drives were physically missing from the Health Net data center it runs in Rancho Cordova, CA. The drives contained personal and health data on 1.9 million of Health Net’s six million customers.

We’ve grown so unsensitized to these data losses that 1.9 million doesn’t seem a very big number anymore. And this particular data loss, since it doesn’t involve some invisible hand reaching through the wire, seems somehow less invasive. That surely must have been the way Health Net felt about it, given this particularly callous sentence from their press release about the loss: “While the investigation continues, Health Net has made the decision out of an abundance of caution to notify the individuals whose information is on the drives.”

Doesn’t this imply that Health Net believes that informing us of the loss of our medical data is optional?

Time for all you HIPAA lawyers out there to tell us what right we have to know when our personal health data has been stolen. Was Health Net just trying to spin this story in a smarmy direction or do they actually have no obligation to tell us?

As for IBM, this loss happened on their watch so what did they do about it? HealthNet outsourced its IT to IBM. IBM outsourcing involves a long check list of things to do to each server to lock it down and make it easier to support. IBM techs install support tools like antivirus and backup. Since they inherit network and application designs from the customer, IBM doesn’t guarantee they are hack proof.

Did you know that? I didn’t.

IBM tries to find problems, I’m told, bring them to everyone’s attention and they try to fix them. Sometimes a problem can’t be fixed or won’t be fixed in which case IBM writes a “risk letter” documenting Big Blue’s concerns and the business risks to the customer.

That’s what is supposed to happen. What really happens is usually a bit different. These days most IBM contracts are under funded to the point of being irresponsible. There may not be time or funding to do basics like securing the servers. With offshoring on top of outsourcing, very inexperienced people in foreign locations are doing much of the support work remotely.

But you can’t blame the physical theft of nine disk drives in Rancho Cordova on an entry-level support guy in Pakistan. This story appeared in Computerworld back in March and then quickly disappeared. I’d like to know what the Hell happened? Wouldn’t you?

As far as I can tell IBM never said a word on the subject.