The template has been stable for over 15 minutes, and the upstream hasn't broken in at least a week. Time to put up a guest password and see what happens.

Thanks for visiting, guys! Look around and tell me what you think, either form or content. Some of it's pretty geeky, some is more general. Sorry if the HTML is a bit lame, I'm spending more time on the content.
11:03:36 PM    

This evening I went over to Stanford to attend a talk by Roger Marks, chair of the IEEE 802.16 wireless MAN committee. Now if this was a venture community event, if would cost you something betwen $60 and $200, though the munchies would be high class. But in this case, the price was $0, which included pizza and pop. I love the IEEE. Herewith a few notes on 802.16/a gleaned from the presentation:

802.16 is for (primarily licensed) spectrum above 10GHz; the 16a variant is for use below 10GHz, and will include the MMDS licensed bands and U-NII unlicensed band (the one 802.11a also uses). The 16a variety will provide non-line-of-site (NLOS) capabilities. Both share a common MAC, which is largely TDM based and designed to provide a variety of quality of service options - committed bit rate, best efforts, polling services. The DOCSIS and carrier inspiration really shows here.

802.16 is doing some really interesting things with dynamic adaption to the environment. It's capable of interspersing different PHY modulations, e.g., QPSK or 64-QAM within the same TDM interval. The less aggressive modulation can be used for more distant stations and the higher bitrate scheme for nearer. The modulation can be backed down dynamically if environmental factors change, e.g., a rain storm. Each frame starts with QPSK and proceeds on up to 64-QAM within the same frame, so the distant stations can just give up and come back at the next start time when they lose the lock. Cute.

802.16a has three (count 'em, three) different PHY options: an OFDM with 256 subcarriers, OFDMA with 2048 subcarriers which are also used as a multi-accesss method, and finally a single carrier option for vendors that think they can beat multipath problems in this mode. The big supporter of OFDMA seems to be Runcom of Israel, based on net gossip. It wasn't clear who's backing single carrier. The weight seems to be behind the OFDM option in the IEEE group, and it's the sole option being supported in the European ETSI parallel standard. The BWIF group seems to have unofficially shut down their competing proprietary OFDM effort. Advantage 802.16a OFDM PHY.

Finally, a mesh option has been slipped into the standard. This wasn't described in any detail, but was originally included (along with a dynamic frequency selection capability) to allow overlapping, ad hoc networks in the unlicensed spectrum. At the last minute, the mesh capability was expanded into the licensed bands as well. Interesting. Wonder what, if anything, the licensed spectrum carriers will do with that.

Overall a very powerful, but also complex, standard. It comes with many of the capabilities that others are trying to retrofit into 802.11 based gear to turn them into impromptu meshes and neighborhood systems. It also carries a lot of baggage from requirements of traditional carriers. How much will that cost in terms of time to market and final product costs?

Oh yeah, some slides were also shown from a presentation given by an Intel Capital person, Sriram Viswanathan, at the WCA conference. It called 802.16a the 'next 802.11'....
10:39:31 PM    

We are at war. Our enemy has a particularly callous disregard for human life. It deliberately attacks civilians and is willing to expend its own people in suicide missions. While those who think America is unwilling to take casualties are wrong, we don't want to take losses if it can be avoided, and we are willing to throw technology and money at the problem. For America, part of the answer will be robots. They are going to get the nastiest and most dangerous military jobs. We will match suicide bombers with suicide machines.

Some first-generation military robots are well know already. The Predator drone aircraft was used heavily in Afghanistan, and has now been armed with Hellfire missiles. Less well known is the Global Hawk, a larger high-altitude reconnaissance drone with inter-continental range . Robots are also in service on the ground. The'Packbot,' a man-portable remotely controlled machine, was taken straight from a DARPA R&D program and sent into caves in Afghanistan ahead of the troops.

There's more on the way. Here's the PDF text of a presentation I saw at DARPATech 2002, describing a micro-aerial vehicle, a semi-autonomous robotic flyer about a foot in diameter, that can float ahead of troops, perch on top of buildings, and send back video of what's out ahead. The Navy is working on Unmanned Underwater Vehicles. The Air Force is prototyping unmanned combat aerial vehicles - UCAVs - meant for penetrating and destroying anti-aircraft defenses.

Reconnaissance, 'walking point', standing sentry, initial penetration of defenses are the nastier end of the military mission. They combine high risk and a need for unblinking concentration, and are best left to machines. Battlebots are coming off the TV screen and going to war.
5:57:52 PM    

So Steve Case is all but out the door at AOL Time Warner, and the post mortem is well under way. Since I diss'ed the deal when it was consummated, why not pile on now?

How did this one go wrong? Let me count the ways:

1. Internet access, and broadband access in particular, is not synergistic with television. In fact, it cannibalizes it. In consumer households, the struggle is for time (attention) and the incremental dollar. Even before broadband, time and money spent on Internet accesss came almost one-for-one from: (drum roll......) cable TV. Early returns from broadband households seem to show that the effect is even greater, with the biggest new competitive factor being MP3 and other downloading. If you wanted to hedge a business against this issue, then AOL Time Warner might have made some sense. But that's not at all how it was sold, or apparently managed.

2. Content is not king. That hackneyed phrase can usually be traced back to - surprise! - content owners. It's been BS since the dawn of online. Dial the clock back to the late '80s or early '90s, the beginning of what we now know was the online and Internet explosion. At that point an online provider needed content, but as part of a bait and switch marketing pattern. If you can get your head back into those times, you may recall the CompuServe, Prodigy and American Online ads talking a lot about the value of online for things like stocks, homework, sports scores and on and on. Yet, the online services already knew that the behavior most predictive of a long term loyal customer was not content access, but rather adopting e-mail, instant messaging (CB back then), and reading and posting on bulletin boards. We were really selling our users to each other.Why wasn't that used as the marketing proposition? Because explaining the benefit is very difficult if the customer hasn't been there already.

Sometime about 1995, that pattern changed rather abruptly. Rather than having to sell the value proposition, the ISPs and online services had customers lining up at the door saying 'give me some of that Internet stuff' - the marketing pattern shifted abruptly away from content as bait. Those who stuck with the old content driven marketing pattern when launching new businesses died ugly deaths. Anyone remember Pathfinder? CompuServe's WOW? I didn't think so.

So repeat after me: Communications is king. Content is bait. Commerce is an afterthought. That's the real value prop, always has been.

3. The instincts of a broadcast content provider are 180 degrees wrong in the Internet medium. The differential power of the Internet as a medium is to support activities such as microcasting, niche interests, and self-organized communities. Everything from cheap tools to unbundled hosting to low production value expectations supports the idea of low-rent, self-produced content that incrementally drags attention away from the mass produced stuff. Why the heck are you reading this blog rather than giving your attention to the WSJ or some other reputable, edited business or technology publication?

The instincts of a broadcast medium owner are to crank up the production values (costs) and hit as large an audience as possible. Dead wrong in this medium. The best they can do is hedge themselves against audience flight by setting up online commerce and fan presences, but even that may lead formerly happy Survivor watchers into a medium where they discover they'd rather read blogs about bunny rabbits and smart bombs. Steve Case knew about this once: the early QuantumLink / America Online grew by catering to communities such as gays and the elderly. Somewhere along the line he forgot it.

4. AOL itself is a maturing business. Since the Internet flip in 1995, AOL has positioned itself as the civilized version of the 'net. Training wheels, if you would. Problem is, a lot of us are already well trained and ready to ride on our own. Think about the penetration level of Internet in the US today. We are already well into the late adopters at the selling edge of the market. Selling in this category is social, by word of mouth from trusted peers who have already had a happy experience. And what single capability of the Internet leads to social adoption? Communications. Mostly e-mail and increasingly instant messaging. Marketing like Earthlink's is the future of branded Internet: We get you on the 'net and get out of your way. AOL would have a tough time keeping growth and share in this situation regardless, due to their perceived position. Shackled to a content company it will never get there.

So other than told-you-so, what's the point of this little screed? When you hear investment touts talking about compelling content, synergy and convergence, run like hell the other way, someone is about to lose their money. This time, it was Time Warner's former shareholders. Don't let it be you. Should be obvious by now? Maybe not to the writers of that Yahoo story I linked:

But the rollout of broadband, and other step-by-step developments, suggest that the digital future of media proclaimed by Mr. Case and others will come at the pace that most technologies advance into society as an evolution, rather than a revolution. This common-sense perspective, to be sure, was not so common a few years ago, before the dot-com crash.

11:23:57 AM