baz.com - Brett Thomas' Blog

I enjoyed xkcd's map of the web, this morning:

Used under general license terms available, there. Click through for a larger one. If you're anywhere as geeky as I, xkcd's great, and highly recommended.

This MSNBC story (linked to by Instapundit) details how researchers are starting to learn, at a cellular level, what happens when people die. And, unsurprisingly, the conventional wisdom was wrong. Everyone figured, enough time without oxygen, and cells start dying. Despite "wanting" to continue living, their fundamental chemical processes stop functioning - "like a candle flickering out when you cover it with a glass" - and that, what's most important is to get more oxygen to the cells as quickly as possible. That's why, when you have a heart attack, they work hard to try to keep your heart and lungs working as much as possible (via CPR). The theory was, that extra bit of oxygen would keep your cells alive, much as a smoldering fire might be kept going with insufficient air, but a completely oxygen-starved fire will extinguish. One has to wonder to what extent centuries-old metaphors between life and fire influenced this view?

Now that they're actually beginning to research the particulars, though, the reality is much different. It turns out, if you starve cells for oxygen, they live fine, for at least an hour. In fact, Dr. Lance Becker, who was conducting the research, was so surprised by this that he noted, "we thought we'd done something wrong." As Asimov said, "The most exciting phrase to hear in science, the one that heralds new discoveries, is not 'Eureka!' (I've found it!), but 'That's funny...'" So many discoveries are made by scientists who do something, knowing full well what the result will be, who then end up spending a lot of time verifying their methods when the result is unexpected. Most of the time, the methodology is to blame, but sometimes, sometimes...

The best theory now is that the cells are fine without any oxygen, for quite some time. However, the mitochondria in the cells are confusing the jolt of oxygen after resuscitation with some sort of cancer event, and are causing the cells to undergo apoptosis (programmed cell death)^* when oxygen is resumed. Those cells then die. In other words, other than being dead, you're fine at a cellular level while you're getting CPR, traveling in the ambulance, and arriving at the hospital. It's when they defibrillate you at the hospital and jolt you with oxygen that your cells then spontaneously undergo apoptosis, and your death becomes permanent.

It would seem, then, that recent research indicating that fast defibrillation is the key to heart attack survival has more to do with avoiding that later flooding of "dead" cells with oxygen than with any inherent short time limit. It may be that, in fact, if we can come up with some slower techniques, heart attack victims can be revived tens (or maybe even hundreds) of minutes after they "die" without serious long-term effects. In this view, the inevitable brain damage from long-term "oxygen starvation" during death is really caused by sudden oxygen resumption after resuscitation.

Dr. Becker also thinks this is all tangled up in hypothermia, too - the fact that people (especially young people) who drown in very cold conditions (e.g., by falling through thin ice on a pond) can sometimes be resuscitated hours later with no ill effects. Perhaps the cold retards whatever initial chemical "priming" is setting up the later apoptosis triggered by the resumption of oxygen?

In any event, all of this is very exciting stuff. It will no doubt take at least a decade before new techniques can be perfected, tried, and become mainstream practice. For the first time, though, we have a view to how death from sudden heart attack can be prevented, and the beginnings of an understanding of how death works at the cellular level.

This is an umitigated disaster. These communiques from the front are the most optimistic and authentic view we have of what's going on over there, and some concerned bureaucrats are cutting that off. No wonder we can't sustain a war when the people in charge are so absolutely clueless about the media war. The other side relentlessly manipulates our media to destroy support in our population, and the Pentagon actively stands down its only ability to respond.

As the update to the article notes - anonymous, disgruntled military will continue blogging. The only ones you're cutting off with this are people who are proud of the job they're doing and play by the rules. Absolutely incredible. Heads should roll in Washington over this, but, of course, they won't.

If you're not familiar with What's My Line, it's an old show in which a guest is asked questions about his "line" (of work). The questions are yes/no questions, and the panel of judges can ask essentially unlimited questions that result in the answer "yes", but a limited number that result in the answer "no". This clip is of Salvador Dali's appearance on the program. It starts slow, but gets very surreal as he insists on answering every question in the affirmative and the in-studio audience begins to crack up:

Via Boing Boing

The Washington Post a few days ago had an article talking about mankind's first real working relationships with robots, charmingly titled, Bots on the Ground. While the last half of the article is irrelevant to the main point - discussing an ROV pilot and an ad for an in-car navigation system - and one of the two main stories is the recycled story of "Scooby Doo" the packbot, from a year ago - I found the description of the centipedal mine-seeking robot interesting. Designed to be self-adapting to its changing shape as mines blow off its legs, the pathetic sight of the battered 'bot dragging itself steadfastly forward on its last remaining leg was enough to get the colonel in charge of the test to halt it because it was inhumane.

It's interesting that in the science fiction that first envisioned people interacting with robots - robots who were much more human-shaped than current models - often envisioned people treating their mechanical helpers callously, until the robots eventually earn (or take) equal rights. Asimov's classic Bicentennial Man is about just this topic. However, it turns out, human beings are so good at anthropomorphizing that we in fact place souls in machines far to simple to posses them. Ironically, perhaps because these robots look like cute little pets, people are as attached to them as, well, people.

The following video clip is from the IMAX movie, Striaght Up: Helicopters in Action. It is of a worker who works on high voltage (750k voltsish) wires, while they're live. The commentary is interesting, but I thought it was fascinating how the whole basic system works. He flies in on a helicopter (so he's not touching the ground). First, he needs to bring himself and the helicopter up to the voltage potential of the uninsulated wires, and the biggest issue to that is controlling the arcing. If you watch closely, he has a rod, which is attached to the 'chopper by a cable. He holds it out, and the wires arc across the air to it as the pilot brings them in, close. The technician then makes contact with the rod, which lets him clip on a lead and temporarily connect the helicopter on the wires. He then carefully transfers his gear, climbs over to the wires, clips his rod to the wires, places the rod on the helicopter platform, and unclips the lead. Finally, he holds the rod out away from him as the helicopter departs, followed by a giant arc of electricity.

It's obvious watching it how consciously pre-thought every single move is. One minor lapse in judgement, one quick loss in focus, and you briefly have 750,000 volts going through you - as you're being blasted by the power off your platform and plummeting several stories to the ground. As long as you keep aware of what you're doing, though, it's perfectly safe!

It's thirty minutes, but this is an interesting video from the BBC about Scientology. The cult recently tried to preempt this by releasing only the video of the reporter losing it, but the BBC did a very nice job of including that in this release. The church was trying to control things, but, in fact, it seems the BBC was even more nimble:

I stumbled on the blog Cosmic Variance which is about...well, I'm not quite sure what. I've read three posts and they were all pretty out there. The last was most interesting - it talks about a bunch of things, not least of them being whether it's more likely that we're the beings we think we are or that we're random fluctuations in the vacuum (or maybe the less-than vacuum).

Really, it's quite beyond summary and I recommend you check it out. Especially if you've been doing drugs...

I'll be giving a presentation at PLUG on Thursday. I'll be talking about building scalable, robust web and application services using open-source technologies, some of the unexpected challenges from that, and my experiences meeting them. Head to their website for specific information about time and place.

This all started as a five-minute extemporaneous rant I gave Bernard Golden, which somehow got morphed into an "oh, about ninety minutes" talk in front of PLUG on Thursday. I'm most interested in challenging people's ideas about a number of bits of conventional wisdom about how to manage the pieces - in particular, databases and third party applications. A lot of people are still using the tools they learned in the 80s and early 90s, and they just don't work, anymore.

I love being enough of a cultural esotericist to know that I am supposed to now throw some over my shoulder and enough of a rationalist to know that it doesn't matter, and not do so.

Freakonomics Blog wonders, in utterly failing to resolve the "is QWERTY an example of how markets don't work or a justly rewarded innovation?" question, "Is there anyone who cares enough about a few extra words-per-minute who would be willing to give up the ability to type for a whole month?"

My personal answer is, "not me." In 1997 or so, I was working for PGP, Inc. At that time, we were working on a new product, called PGPkeyserver, which would be a repository for people's public keys. I kept having to reference it in documentation and the like, so I had to type it a lot.

My typing is particularly idiosyncratic. I learned to type by posting and chatting on BBSs in the 1980s, and cemented that basic knowledge to touch typing in the early nineties by sitting in dark rooms, coding all night (I love the dark). I do about 65 words per minute (or did, the last time I was tested, which was about fifteen years ago), but I use four fingers on my right, dominant hand (all but the ring, though I only use my little finger for hitting "enter") and one finger on my left hand - my index finger. As you might imagine, my left hand flies all over the keyboard as my left index finger tries to do the work of three on my right hand, and fast enough that the right hand doesn't get ahead of it. More than one person has commented that watching me type is like watching someone play piano, since my left hand is flying all over the place.

I suspect this methodology may help explain my relative success in avoiding carpal tunnel syndrome - I don't just keep my hands in one place, they're constantly moving, so I'm not really making the same moves over and over again the way conventionally trained typists are. It also has a neat side effect that I can type one handed, or even when the keyboard is upside down to me, because my brain knows "where the keys are" and not "how to move each finger relative to the home position." My home position moves around - my left hand often rests with my middle finger on "2" and the index on "e," I guess because "e" is probably the character my left hand hits the most. Sometimes, though, the index finger is resting on "g." And my right will cross the middle - when I type "character," for example, my right hand usually gets the "t" and sometimes the "r". How words gets typed often depends on where the hands are from the previous word, so it's very dynamic.

My "method" - such as it is, as it's what came from just typing lots and lots - works pretty well for most words. But, if I hit a long word where every letter in it is on the left hand, it's really jarring and really slows me down as I suddenly have to peck out every individual key with my left index finger. "Keyserver" (or worse, "keyservers") is such a word - I get the "k" and the "y" with my right hand, but it's like hitting a speed bump for the rest of it, with all the keys on my left hand.

I finally decided, back when I was writing that documentation for PGP, "alright, this is stupid." Dvorak is supposed to be designed so that it's more balanced. All the vowels are on the left, there, and since vowels and consonants tend to alternate in words - and, to the extent they don't, it's usually consonants following constants, or the same vowel doubled - it would seem I'd be especially good for Dvorak. I changed my computer over and took the plunge.

I believe I made it about a week. For the first few days I was OK - I think I started on a weekend, to give myself a head start. But, quickly, I hit the point where I had to do a lot of work, and it was just over. It wouldn't surprise me if I could hit 75 or 80 words a minute with Dvorak, but that week was just awful. Typing for me is so automatic - I think the words, and they appear on the screen, it's really very close to a telepathic interface. I don't think about it at a low level, at all (except when doing something like this where I'm thinking about it because I'm writing about it). To suddenly have to think about every letter I typed was agonizing. My first thought, at the time, was that it was like being unable to speak, but I realized it was worse than that - it was like being unable to think. Where I usually am thinking at a high level about either what I want to write or what the program I'm working on should do, when I tried to learn Dvorak, I had to think about typing, instead. I had a lot of trouble thinking about anything else at the same time.

Personally, I'm done. It's definitely not worth the cost of switching. Do I wish I'd learned Dvorak at first, not QWERTY? Maybe, I guess. Mostly the experience made me appreciate how cool it is to integrate your brain with a physical device so that it becomes more of an extension. Kind of like those guys who drive a piece of construction equipment for twenty years and moves it so smoothly, it's like a part of his body.