connellyenterprises.com

My last column generated a lively debate on the prospects for various business and technical options for the delivery of Internet TV so it makes sense to continue this topic and build it into a more full-featured model.  I used to write quite a bit about this back when I was trying to get NerdTV going.  The core of what I’ll write here can be found in a couple dozen columns from back then — columns that would seem to have been for the most part forgotten given the direction last week’s discussion took.  You see the future of television IS Internet television.  There is no other in sight.

No business or technology exists in a vacuum.  They all have customers, users, competitors, and make use of resources in an environment that is not one of total abundance.  This means that if there is going to be something like television in the future it is going to adapt to the distribution model that offers the highest price/performance, which is to say the highest performance for the lowest cost.  That is not how one would traditionally describe the Internet, but then times are changing.

Whatever country you live in there are generally four models for live entertainment video distribution — broadcast, cable, satellite, and Internet.

Broadcast is a limited local resource and therefore more highly regulated than the others but it has traditionally featured the lowest cost per marginal user.  That means it costs a lot to build and maintain a TV station but additional viewers within the service area can be added pretty much for free.

Cable offers more channel capacity than does broadcast but requires building a distribution network that’s fairly expensive.  While one could imagine a cable TV “station,” the way the industry has grown is through cable operators becoming content aggregators offering many services over their expensive networks.  That’s the most efficient way for cable companies to serve the broadest audience and the only way that enables them to sell extra-cost services like pay-per-view, premium movie channels or, indeed, Internet service.  Remember, though, that cable operators pay for nearly all of the content they carry, which is different from broadcast, where a lot of content is free to the broadcaster and some content even comes with money attached.

Satellite operators pay for their content, too.  Satellite initially used wireless technology to offer cable content in rural areas where it was too expensive to build a wired network.  Having gained economies of scale in the rural markets cable couldn’t compete for, satellite has come to town competing generally on price.  But satellite offers no practical Internet service.  I know there are some and I tried one years ago (Starband) but they don’t work well.

Internet TV is different from all these others.  It began as a parasite on telephone and cable networks so the cost of building the network generally wasn’t there, having already been covered for the most part by those earlier services.  Internet TV is less of a network than a conduit; at present the Internet Service Providers don’t pay for video content but then neither do they get paid for it.  Yet this common carrier attribute also makes Internet service often more profitable for telcos and cable companies than the core services those companies were established to provide.  Whatever you pay for Internet service, it is mainly profit for your ISP.

The important lesson to learn when it comes to these competitive services is that the first three — broadcast, cable, and satellite — are all going up in cost to their providers while the cost of providing Internet service is going down.  In the USA, broadcast viewership is dropping, which means the cost per viewer is rising.  Same for cable where viewers are stagnant, viewership is declining (number of hours of viewing) and the cost of content is rising.  Satellite has been growing marginally but that could end at any moment and it shares the same content cost increases as cable.  Meanwhile Internet service just gets faster and cheaper thanks to a Moore’s Law double whammy.

Remember Moore’s Law works in two ways.  It makes digital products ever cheaper AND ever more powerful.  This has profound meaning for Internet TV because it continually increases the bandwidth we can get for the same dollar while giving our devices the capability to do even more with the same bandwidth.

Here’s an example.  My primary Internet connection is an 8 megabit-per-second business cable line with a service level agreement and static IP addresses.  I pay more than you do but then I get more, too, though even my service is crap compared to what you can get in Japan, Korea, and much of Europe.  My primary computer WAS a Mac Pro G5/1.6 circa 2004.  I should have replaced the G5 a couple years ago, I know, but my kids are in private schools and I keep buying airplane parts. I finally replaced the G5 last week, though, with a dual-core Mac Mini 2.0.  Both the old and new computers had four gigs of RAM.  Though my Internet connection can easily carry one or more 1080p H.264 video streams, there is no way that old G5 (which cost me $1999 in 2004 dollars) could play it.  It didn’t do much better with 720p for that matter.  But the $750 Mini (small drive but lots of RAM) can easily decode 1080p.

This is the trend, then: our available bandwidth will go up while our devices will become more powerful, making better use of the bandwidth.  The result, as always with Moore’s Law, is either better services or lower total cost or maybe a little of both.

What this means for the future of television is that we’re approaching a point where Internet service will equal and then be lower than the marginal per-viewer cost of the broadcast TV model.  This crossover will inevitably happen with the only question being when. That’s a function of bandwidth costs decreasing at 50 percent per year and processing power increasing at 50 percent per year.  My calculations suggest the crossover will happen around 2015, which used to seem like a long time away but no longer does.

When Internet TV becomes dramatically, unequivocally, and inexorably cheaper than the other three distribution models, those other models will quickly go away.  That’s why I argued in PBS meetings to forget about spending $1.8 billion to upgrade local stations for digital TV and instead sell or lease that spectrum for commercial data use and throw the resulting $3 billion (lease revenue plus the $1.8 billion savings) into rebuilding the network solely as an Internet service.

Nobody listened.

So there is a cliff rapidly approaching for television.  Five years from now local TV stations will have the same complaints that local newspapers have today as many of them go out of business.  Cable TV operators will become ISPs, period.  Phone companies will be ISPs, too, and analog voice service will be gone completely.  The regulatory implications of these changes should be interesting.

Who, then, will be the players in this future TV?  For the most part they will be the content providers, which probably doesn’t mean traditional networks.  And the networks know this, by the way. Hulu.com isn’t called NBCFoxABC.com and TV.com isn’t called cbs.com for a reason. Networks will go away.

But content will endure, bringing new value to I Love Lucy episodes and almost anything else people like to watch.

The TV networks are throwing their lot together.  CBS chairman Sumner Redstone will come to his senses one day and merge tv.com into Hulu, I am sure.  Their big competitors will be Google, Apple, and a player yet to be even founded (definitely NOT Yahoo OR Microsoft).

Google will differentiate itself as always through technology.  Those shipping container data centers I first wrote about in 2005 exist not just because they are easy to stack inside big Google plants.  Why botehr with weatherproof containers if they are to be used exclusive indoors? Because they are even easier to put in the parking lot at the telephone company central office or at the cable company head-end, both of which will by then be strictly ISPs.  Google will proxy content at every major ISP in America.  And they’ll do this because Google has no idea what people want to watch on TV, nor do they particularly care.

Apple, on the other hand, cares.  Following the content development scheme I laid out last time Apple will attempt to become the dominant content provider to the 20 percent of the market that spends 80 percent of the money, with margins high enough to use Google distribution and still come out ahead, leaving to Page and Brin the 80 percent of content that generates 20 percent of revenue.

But wait, isn’t Apple just a maker of hardware?  Don’t they do iTunes just to sell iPods?

No.

Apple is a software company that has traditionally packaged its software in attractive hardware boxes.  The fact that any new Mac is essentially a Windows computer proves that.  But price points have been eroding in every hardware category and will continue to do so.  Microsoft right now makes more profit from every Windows PC than does the maker of that PC.  Apple is not immune to this trend.  So the company needs to find ways to sell more and more software.

Content is software.  TV is software.  And the great thing about entertainment is that it is software we can be induced under some circumstances to buy over and over again like those teenage girls who paid to see Titanic dozens of times.

What does that leave, then, for that player to be named later?  I’ll get to that next time.

Not long ago I had a chance to visit the big data center at 365 Main Street in San Francisco.  I was invited by friends to help them install the first servers for their startup, which is still in stealth mode.  The data center was enormous, though my friends occupied only a small part of one rack not far from the Oakland Raiders and one floor up from Bebo, the social network bought not long ago by AOL.  Bebo is a big hit in the UK and I found it odd that all those British profiles are hosted in San Francisco, eight time zones away.

We installed the servers — three little boxes and one big one — then fired them up.  In moments the site was live, though with only a few alpha clients.  The application, which I am sworn not to mention yet, is clever, but not particularly resource intensive.  It’s just like any other web site only a little different in several good ways.  The three little web servers just sat there blinking, doing their jobs.  But the bigger box (a 3u, versus the 1u web servers, if you are keeping track) was wailing right from the moment of booting.  Inside were three 15,000-rpm drives, a bunch of processor cores, and a ton of RAM — all of it thrashing away despite the small load.  What was going on?

What was going on was the twilight of enterprise application development as we know it today.  That 3u box was a database server that sits behind the three little web servers and makes this new enterprise work.  And work hard, apparently, for all the drive thrashing and lights blinking.

We’re at an interesting point in the development of computer technology.  Processors, having been failed somewhat by Moore’s Law in their attempt to become more powerful by widening data paths and raising clock speeds alone, have now resumed or even accelerated their performance growth by replacing one processor core  with 2, 4, 8, and eventually hundreds of cores, most of them not really needed.

Processing power isn’t what binds enterprise or Internet applications today.  I/O and disk access do that.  Servers have one or two gigabit Ethernet connections, each of which could be easily saturated by an old Pentium 4.  It’s the pipe that limits us, not our ability to pump bits through that pipe.  Thanks to the gamers, I suppose, and to a surreal and not particularly useful competition between Intel and AMD the main server CPUs are barely sweating even though they are running the core business logic of the application.  It’s the database server with its disk drives that is working so hard, grabbing data to feed the web servers seemingly just in time.  But don’t blame the hardware here or even the disk drives — blame the database.

We’re at the apex of SQL database development.  It’s 1890 and we make the best darned database buggy whips on Earth.

There is a better way to handle large volumes of data and that better way has been established, not surprisingly, by Google with its BigTable semi-structured database that essentially caches the entire Internet.  HBase from Hadoop is the Open Source version of BigTable and both are rapidly making old SQL databases like Oracle and DB2 obsolete for certain users.

Amazon.com runs on an Oracle database, but one that was extended and optimized at a cost of more than $150 million.  Amazon probably represents the most that one can do with SQL in terms of scalability.  Anything bigger requires a completely new approach like BigTable.

Or maybe it isn’t so new at all.  I recall something very analogous to BigTable during the network operating system wars of the 1980s.  Microsoft had a couple dozen OEMs working on network operating systems based on the hierarchical file system of DOS 2.0 (Paul Allen’s last technical contribution to Microsoft).  While a hierarchical file system may have made some sense for a workstation it made little to no sense for a server accessed by dozens of workstations in the view of the programmers at Novell, where Netware was being born at the time.  Those guys ignored the hierarchy and wrote the entire File Allocation Table for each drive to memory as a single flat file called an Indexed Turbo FAT.  Where the DOS-based network operating systems had to search the disk for files, Netware had the entire index loaded in memory and instantly knew where the target data could be found.  The system was easily 100 times as fast.  BigTable takes this a step further, I suppose, by ignoring the distinction between index and data, dramatically expanding the memory footprint but, at the same time, completely eliminating a retrieval step.

An irony of BigTable and Indexed Turbo FATs is that both Google and Novell were pretty upfront about what they were doing and why, yet competitors have remained bound to lower performing technologies because, well just because.

Which brings us back once again to Oracle buying Sun, a deal that has continued to bug me because it didn’t make sense… UNTIL I thought about it in terms of the scalability of SQL architectures and market positioning.

Right now almost every web application has an Apache server fronting a database box running MySQL or its closed source equivalent like Oracle, DB2, or SQL Server.  The data bottleneck in all those applications is the SQL box, which is generally doing a very simple job in a very complex manner that made total sense for minicomputers in 1975 but doesn’t make as much sense today.  Five years from now the situation will be very different with HBase running everywhere, the dedicated SQL box eliminated completely, and the database shared across redundant web servers like a micro-Google.

Where does this leave Oracle?

It leaves Oracle bleeding its big stupid corporate customers for another decade but eventually losing both the bottom half of the market and the very top where applications scale to tens of thousands of servers.

Part of the distinction here is between running a mobile phone billing system in one case and Facebook in another.  In the mobile phone example you’d better get all those minutes or money will be lost.  But in the Facebook example reality is more approximate and if an update propagates slower than expected, well big deal, so you missed Little Johnny’s birthday pictures for an extra 20 seconds.  There are even business software cases where this philosophy applies.  Progressive Insurance, for example, is always ready to give you a comparison price quote for auto insurance not because they can generate that quote (and the price quotes of their major competitors) on the fly, but because THEY GENERATE A SPECULATIVE PRICE QUOTE FOR EVERY CAR IN AMERICA EVERY NIGHT.  They don’t generate a quote when you call, they just access it because it is already done.

So Oracle keeps the mobile phone company as a customer but doesn’t keep Progressive in this example.  And in the long run there’s enough data redundancy built into the loosey-goosey HBase model that it becomes just as reliable as the more rigorous SQL model that it is inexorably replacing.  That’s when Oracle loses the mobile phone company, too.

Larry Ellison won’t like that.

So what’s to be done?  Buy Sun.  Get into the database appliance business.  Start selling highly-tuned database appliances that achieve the simultaneous goals of vertical integration (making profit on the hardware as well as the software), obfuscation (keeping the customers out of the lower-level code by encasing it in an appliance), and increased overall performance (putting off the inevitable loss of market dominance for another three years through a hardware tour du force).

IBM, as the other big SQL company, doesn’t really share Oracle’s problem, because IBM makes money from the hardware already.  If DB2 gives way to something like HBase, IBM will run HBase on its premium iron — a luxury Oracle can’t share without buying Sun.

As hardware gets cheaper we extend performance by distributing software across more and more machines.  But that distribution in itself undermines the lucrative software licensing system.  So we introduce a new level of abstraction — the database appliance.  Prices will go up a little while performance will go up a lot. Customers will think they are getting more for their money and they will be. But the ultimate comparison that has been at least postponed is between paid and free, where free always wins in the end.

And THAT’s why Oracle NEEDS Sun — to extend its current run by another three years, buying Larry time to write an Act II for his company.