Sunday, October 26, 2008

UWB and the sins of another monopolist

There's a little history I want to get down before I forget too much of it. It has to do with wireless Ultra WideBand, used as a local interconnect technology.

The wikipedia (and other) articles have already been so denatured, lost, hidden, covered up, re-analyzed, etc., that the cuplrit seems a saint, and the tragic hero seems nothing but a buffoon.

I have no specific evidence. Motorola was, as large corporations tend to be, depending on secrecy as a weapon, hoping to get a leg up on iNTEL by surprising the market with fait acompli devices that would become defacto standards before iNTEL could foist its junk on the world yet again.

Uhm, no, secrecy is not a good weapon. So the tragic hero is something of a buffoon. Oh, well.

So my brother, who was directly involved, did not encourage me to do anything that would have blown their secrets, and I did not record things as I should have. So I can only give a rough outline of the events.

Somewhere around the year 2000, while everyone else was worried about the possible effects of cutting corners in business computing on dates, and on mechanized control procedures that depended on dates, my brother was arranging a soft landing as he was processed out of Motorola. A couple of friends who also left about the same time had started a company, eXtreme Spectrum (or something like that, remember the X fetishes about that time?) to work on utilizing certain "under-utilized" radio bands for low power, non-obtrusive local interconnect. Those friends invited my brother to join them, and he did.

The technology borrowed heavily from spread spectrum cryptography techniques. (Think Hedy Lamarr. Think no data wires between your computer and your printer and your monitor and your disk drive and your speakers. Safely. No spilling the data out into your neighbor's living room or the war-drivers' laptops.)

There was a rash of companies that were competing to bring a UWB technology to use that spectrum before the IEEE and the FCC at that time, but, according to my brother (and according to my own research), only eXtreme Spectrum was using the spread spectrum techniques right.

Because of the cryptographic implications, I was concerned that the NSA and others who think that secrets are powerful weapons would scuttle their efforts. I don't know of any specifics to indicate one way or the other about the involvement of the NSA et. al., FWIW.

I was also worried about iNTEL's (lack of) willingness to play fair, for the obvious reasons.

(Owning a piece of the pipes everyone is using is always a lucrative business plan -- on paper -- and the Constitution be damned.

Why is it that so many people think they have some right to be the only ones not required to play fair? Why is it that they can't see far enough ahead to recognize what that does to their own future business environment? What is the blindness of the guy who wants to be the "benevolent" tyrant? I mean, there is a reason there is only one God. Perfection means everything, and none of us who are not perfect -- be honest, now -- are up to the job. We always, always foul our own water supply. Upstream. Where we're going to have to drink it as it comes down.

No escaping that. You may think you can always jump streams, but you're only fooling yourself. There is only one stream. What goes around comes around.)

iNTEL bribed a lot of companies. (Okay, I have no proof except for silly behavior on the part of said companies.) They also set up a lot of phantom companies. (No? Well, where did all those companies in their forum go?) They played the "future feature list" game. (I. E., "The information presented here are forward looking and depend on technologies we hope to develop in the future, and we tell you this in the fine print so you can't sue us later when it turns out we promised 200% more performance than you're implementation will ever achieve.") The organized a forum of fools. They put out cheap hack "samples". Every classic dirty trick.

eXtreme Spectrum did draw first real silicon, delivering a complete integrated circuit product that was scheduled to be used in some (I think it was Samsung, and other) consumer products. The first chip was capable of streaming two or three clean video feeds across the air in large auditoriums, at low power, without dropouts, and without interference. It got FCC approval.

(My brother said that Apple was talking with them about plans for the iPod. Yeah, that was about two years before Apple "switched". Or twitched. Whatever.)

iNTEL claimed that eXtreme Spectrum somehow broke the rules by going to the FCC to get approval instead of letting them bog the thing down in the IEEE committee. (Oh, yeah. The committee. ISO isn't the only one.)

Motorola eventually bought eXtreme Spectrum a little before the digital semiconductor sector was split off as Freescale, and then there were two players left standing. When it was clear that iNTEL's group was never going to concede the playing field, Motorola or Freescale, I forget which, did not just offer to completely open their tech, but actually did. Not just so-called reasonable and non-discriminatory terms. No IP toll, period.

Of course, I knew that would not satisfy iNTEL, iNTEL wants to own the pipes outright. Being able to participate fairly is not in their plan. Period.

Well, the IEEE committee ultimately caved. The standard efforts were discarded.

And on the bloody morning after, we have wireless USB. On paper. Someday. If you really want to spill your data all over the ether on a "bus" that allows only one master. (Put in layman's terms, think of what it would be like if only one bus -- the other kind -- were allowed on the streets at once, remote controlled by the bus company.)

One tin soldier rides away.