True. But happily, there is data, and the data says that the expected lifespan in Canada is 4 years longer than it is in the US, the infant mortality rate (deaths per 1000) is 5.1 instead of 6.3, yet they spend $2000 less per capita, and instead of spending 15.4% of GDP on healthcare, they spend 9.8%.
There's your data -- the Canadian system is better. Your friend's dead uncle is just an anecdote. Perhaps the surgery he 'needed' was highly unlikely to extend his life.
Most modern countries with universal health care get better outcomes (healthier, lower infant mortality, longer expected lifespan) for less cost per capita and a smaller percentage of their GDP. In their case, it appears that a bigger insurer worked pretty well, plus they don't have anyone going bankrupt from medical bills. If it works for Canada, England, France, Germany, Japan, Australia, Switzerland, Sweden, Israel, Iceland, New Zealand, Italy, Spain, Norway, Greece, Austria, Netherlands, Belgium, Finland, and South Korea, it can probably work for us, too.
(source: http://www.nationmaster.com/ who gets their data from the UN, OECD, CIA Fact Book, WHO, and similar reliable sources).
Cannabis is a good deal less toxic and/or addictive than quite a few over-the-counter drugs. Its side-effects (primarily, intoxication) are common, but not life-threatening. Against painkillers, it compares especially well.
I do agree, however, that if it were marketed as a drug, the guys selling it would find some way to push the truth envelope.
But this is not exactly what charging by the bit does. There are two problems.
#1, charging by the bit does not relieve the immediate congestion. Instead, it creates a customer-annoying surprise at the end.
#2, this is very, very time sensitive, and customers don't have a good way of knowing what is going on at which time. If the network load is under its maximum practical capacity, there's no problem to solve, and charging by the bit is merely a handy way of getting extra money. If the network is at capacity, there is no particular guarantee that the people who are clogging it, are your multi-terabyte P2P downloaders -- perhaps they confine their big slurps to the late evening hours, perhaps a bunch of otherwise light users all decided to see the live Victoria's Secret show. Charging the P2P users for their off-peak consumption, or killing all the P2P connections, doesn't do anything to reduce the bottleneck; in fact, the incentive is completely disconnected from the people causing the problem.
It would be nice, but is not yet practical, to have something like "auction-class" bandwidth. People get (say) $40 worth of micro-tokens to spend on bandwidth, and browsers, P2P, etc, bid for bandwidth. Configure the browser to know roughly what your cost preferences are (based on browsing history) and if you go to visit a movie that exceeds your usual cost, or if you start watching a lot of movies in peak hours, the browser can let you know that you're starting to spend too much "money". Once your monthly "money" is spent, you're stuck at bulk rate. And P2P guys, they would prefer bulk rate in general, so they would make "bulk rate" a pretty unpleasant place to be. This could have its own problems -- ISPs could dodge network neutrality by charging a higher rate to access Google, in hopes of forcing Google to pay some sort of ransom. But, lacking "auction-class", you can keep track of who is the source of whatever traffic is spiking capacity to the limit, and you can (if I understand TCP correctly) slow them down. This might well have the effect of blocking P2P during peak hours, but P2P works just fine off-peak, too.
There's two inconsistent lines of argument here. One is that it is all at the last mile (in which case, it becomes technology dependent. With DSL, as I understand it, the last mile is "my" piece of copper wire), and everything is cheap at the datacenter.
The other claim is that the problem occurs at the datacenter, when everybody wants to watch the bandwidth at the same time. So, which one is it?
My suspicion, not incredibly well informed, is that it has more to do with competing lines of business. After all the whole P2P throttling thing that Comcast did (which they say they are not doing now, though oddly enough, P2P remains broken for me) could have instead been invested in simple peak/non-peak bandwidth-based throttling (i.e., in a protocol-agnostic way, at any bottleneck in your network, penalize the guy(s) who are actually spiking the demand). They chose, instead, to do all this deep packet inspection nonsense. Why that, and not a more direct regulation of delivered bandwidth exactly when it was a problem?
They're coming pretty steadily. I read the FA, logged in to the ssh-receiving box, looked at the logs, and a half-dozen showed up while I was looking. Makes me wish I had the time to run a honeypot.
Based on my quick reading of the ssh/sshd documentation, I created an empty ~/.ssh/rc file for the accounts with known-good passwords (i.e., mine) and then put "kill -9 $PPID" in/etc/sshrc.
That seemed to work, though I know it will be bad for X11 (which I never use over ssh to my home box).
There may be more elegant ways to do this, but if there are, I didn't find them quickly (which is to say, the ssh documentation, like all Unix documentation, sucks rocks).
What I understand is, that LEDs will get cheaper in a "few" years. Apparently there's a little bit of expensive sapphire in each and every power LED, and replacing that is the key, and at least two different research groups have made progress on it. The next hurdle after that is building a good-looking and/or compatible fixture that keeps the LEDs cool enough. Overheat them, their output and their lifetime both go down.
There's also economies of scale that have yet to kick in, and there's the (as we have seen) charlatan factor at work. Enough people will sell crap LEDs to give the good ones a bad name.
The drive to conserve products costs also drives some pretty silly economies. I've been playing with bicycle lights (go look at some prices, be appalled), and if you are willing to pony up for a "generator" (really alternator) hub, you can, for under about $100 in parts, light your bike up like Las Vegas. You can run a standlight, you can charge batteries, you could run a little cell-phone charger. But if you look at what is offered for sale, it's diddly-shit. The "standlight" you get, if anything, is this itty-bitty thing run from a capacitor, not continued operation of your running lights at rated power. Companies are assembling the minimum, and selling it for what the high-end market will bear, which is surprisingly much.
I'll do the best I can here. Short answer is, LEDs are still too expensive for general use, but they already win for certain applications. The three places where I think they win are (1) under cabinets, (2) loading docks, and (3) bicycles.
The best LEDs are now a hair more efficient than fluorescent, but you won't want to use the best LEDs unless you enjoy reading by the light of an arc welder. "Neutral" LEDs are a hair less efficient than fluorescent, but the light is all sent to one plane (assuming you get one with a Lambertian lens on the LED, which is usual for power LEDs, e.g., CREE and Luxeon) which I think works to their advantage. In general, if you need N watts of fluorescent light, you need N, or nearly N, watts of LED light as a replacement.
Conversion power losses, which you mistakenly quote from the headline, are about the same. Cheap wall warts are about as bad as CFLs or cheap fluorescent ballasts.
LEDs are claimed to last 50,000-100,000 hours, and that is not burn-out, but reduced efficiency (70% of rated, I think). Better cooling increases lifetimes; they like to not be hot, and heat sinking is actually a major issue. This is sort of a pricing problem, because the temptation is to economize on the LEDs and up the current (the ones I like for indoor use are rated for 350-700mA, but you don't get double the light at double the current). At the 350mA loading, each 1W LED (bought retail, mounting on hexagonal pucks) costs about $8. 50 watts of LED, which would be a buttload of light, would be $400, plus drivers (you'd need 6 of the ones that I've used, and at $30 apiece, that's another $180). So when I say "expensive", I mean it. You can do better than my estimates, but it would require work, planning, design, etc.
To my eye, LEDs have a better light. There's at least one person here claiming otherwise with great authority, but I think they are working from faulty information or cheap-N-sleazy LEDs.
The reason that LEDs work well for undercabinet lighting is that you don't need to do anything special for them; they are so low profile, you can just tack them up bare. The light goes exactly where it is needed, and that sort of lighting is generally not so cheap so they are more nearly cost-competitive. For loading docks, LEDs are durable, and come on quickly in the cold, and don't need frequent replacement. On bicycles, LEDs are durable enough (CFL on a bouncing bike? I don't think so) and very efficient, and work at a low non-scary voltage.
Poke around at ledsupply.com, they have a good sampling of LED-related stuff.
Depends a whole lot on how cold your "cold" is, and which CFLs you use. We have a couple in ceiling fixtures, that in a Boston winter, get kinda cold, and start pretty slow. They're definitely less than five years old, because the ceiling is less than five years old. But I am pretty sure that these are also cheap bulbs, mostly because there's no assurance that an expensive bulb is actually any better, even though some of them might be. ("We promise our CFL doesn't suck" -- if I were a cheap CFL huckster, I'd slap that label on my crappy merchandise 15 seconds before I tripled the price.)
Exactly true. I checked the power factor on the wall wart running some LEDs with a kill-o-watt, and it was something like.6. The problem is entirely in the wall wart, of course, since the LEDs draw DC.
Most noticeable with Amber -- if you swing the junction temperature from 100 c to 0 C, the light output goes up by a factor of 4.
(So says the Luxeon III data sheet (ds46), page 9).
Re:LED is a viable option in 40 Watt replacement
on
CFLs Causing Utility Woes
·
· Score: 4, Insightful
I think your information is basically wrong. Perhaps you've been buying the wrong lights. The lumen ratings for bulbs take the human visual color response into account, so a "white" LED of 100 lumens is dumping a pretty good load of light into the green range. The color temperature on the most efficient LEDs is a hair off (looks like a welder's torch to me), but their light is as good as what you get from similarly efficient fluorescent bulbs. The LEDs I have in mind are the relatively recent CREE XRE power LEDs, that run at between 350 and 1000 mA.
If you are willing to take a minor efficiency hit, go for a quality "neutral white" LED, and you get pretty good light. I did this for the lights under our kitchen cabinets, and it is great.
You are correct that there are no efficient "direct" green LEDs, but the phosphor downconversion is no different from how it is done in fluorescent lights, and because the initial light is not confined to a small number of wavelengths, the resulting light
is better looking. I initially got this information from a physicist (we were taking our kids on the same ski trip) and I later confirmed it for my own self with a diffraction grating. By-the-way, if you look at those diffraction-grating pictures, you will see that the white LEDs do indeed emit a good amount of yellow and green.
If you see flickering in your LEDs, you are using a truly cheap regulator. All the ones that I use are built around switching power supplies for efficiency, and their "flicker" is in the 10-100 kilohertz range. As far as durability goes, I have generator-driven current regulated LEDs on two of my kids' bikes and two of my bikes, and they ALL live outdoors, one not under any cover, in Massachusetts, year round. The kitchen lights are on almost continuously, since they are the most efficient lights in the house by far. Most important to long lifetime, is adequate heat sinking, and/or not over-driving the lights (they're also more efficient at lower power, even holding the temperature constant, and they're more efficient at lower temperature, which is a side-effect of lower power when you don't hold the temperature constant).
Where the heck did you get your information? If this is from lights that you've bought/used, I'd love to know which ones, so that I will never buy them by accident.
I don't quite understand your analysis -- the reasoning behind Hillman's conclusion is, as I understand it, that non-cyclists (as a group) tend to not get enough exercise, and the risks of THAT (insufficient exercise, as a group) are 10 times higher than the risks of getting killed on a bicycle. To point out that in fact, the crash-death risk of cycling is higher than the crash-death risk of caging does not refute this, unless we are somehow less dead when we die from the heart attack or stroke that is (apparently) much more likely to kill the cager. (And note that Hillman is counting expected years of life lost, not deaths.)
All that said, there are safer and less safe roads, and of course the prudent cyclist will take care to avoid obvious dangers. There's hills I won't descend down (I coast at 35+mph, I measured, and the road is skinny with poor sight lines), and there's roads I avoid in the dark (again, poor sight lines, poor visibility). But note, in all cases, that it is not he bicycle itself that is dangerous -- I avoid the hill because a car might not see me, I avoid the skinny curvy dark road because a car might not see me. The danger comes from the cars, and it would be lovely if more people recognized this.
I'm not sure where this Segway gadget fits in. It still seems to use plenty of energy (35mph, with that big fat profile, is going to use a heap of power, by bicycling standards), and it hardly has a monopoly on being powered by electricity (there's some practical, not-too-expensive, e-assist systems used in cargo biking). It looks heavy, and being wide like that, it will be prone to getting stuck in jams. If the batteries fail, your options are not so good. And it will encourage people to not get the exercise that they need, so that risk (which is apparently enormous) is not reduced.
Which metro area do you prefer, Tampa/St.Pete, Houston, Silicon Valley, or Boston? I ride about 50 miles per week, about 40 of that is on roads shared with cars. But it is true, Boston is well-known for its careful, considerate drivers, so perhaps things are different elsewhere.
The risk estimate is not a silly statement. The estimates are that driving to the exclusion of biking, is 10 to 20 times more dangerous (in terms of expected years of life lost) than biking. This was measured in England (the cycling risks vary from country to country, but so does obesity) by a guy named Mayer Hillman at the Policy Studies Institute at the University of Westminster. Biking in England is less safe than northern Europe, but safer than the United States.
I think you need to do a more rational assessment of risks. There are safe roads and unsafe roads, but overall, not getting enough exercise, is very, very unsafe, long term.
You could move to someplace civilized:-). And yes, I know, some places are just screwed up. Riding in Florida as a kid was a mixed bag; the heat makes you sweaty, but warm rain is no big deal. Timing helps; if there's an oil shortage (as there was then) the jerk fraction goes down nicely.
Walking's relatively dangerous, at least according to statistical studies. Makes no sense to me, but that's what the numbers say.
Bicycles can carry quite a lot of "stuff" -- mine (xtracycle cargo bike) is comfortable up to about 100 lbs, and it's rated for about 200. I do almost all shopping by bike nowadays.
Others -- Long Johns, and Burrows 8 -- are more suited for serious lumpy loads, and put them nice and low.
You're taking your life in your hands when you drive everywhere, too, it just isn't so obvious.
The risks from heart disease and other afflictions of the unfit are much larger that those from cycling,
though perhaps your roads are unusually unsafe.
The weather problems of cycling are also overrated.
The cars are much more dangerous in bad weather, so you have to be extra-careful of them,
but clothing is a big help at dealing with the cold, and lack of clothing helps with the hot.
Rain is, indeed, the worst.
Segways, on the other hand, what happens to them when you ride them through foot-deep water?
How do their batteries cope if they're left out in 20-degree weather? Can I put snow tires on them?
They look at little wide for threading through stopped traffic. And won't all that sitting around make me fat(ter)?
I both ride and drive, both on the road, usually. The health benefits of biking are largest in the first 50 miles per week, and with a 10-mile commute, kids, and weather, biking every day is impractical. When I drive, my first priority is not to hit people, my second priority is to not hit cars, my third priority is to obey the laws. Last on the list is get-there-fast. Even putting speed as the last priority, cars are plenty fast.
And if you don't think that you're competent to drive safely around bicycles (note that the car is the source of the danger, not the couriers), why are you driving? Seriously -- how can you say something as ridiculous as "or I will run you over"? Doesn't your car have a brake pedal? Don't you stop if things look uncertain, as if your car might hurt someone? You're wearing a 2000-pound suit of armor -- it's not like the courier can hurt you. Why aren't you being more careful that you don't hurt someone with your armor suit?
Your choice of domicile has no bearing on the relative safety of 85mph, 45mph, or 15mpg tailgating-induced crashes, so I don't quite see how that is relevant. If you were interested in maximizing your safety (including all risks), you would also get some physical activity several times per week. Speed of travel isn't everything.
And hitting a solid object at 85mph is also "often fatal", and you therefore argue that the risks are equal? Slashdot reasoning at its finest.
Oddly enough, 15mph is a respectable speed on a bicycle, and up to at least 50 miles per week, time spent on a bicycle is more than paid back in extended lifetime from improved fitness, even including the years laws to the rare bicycle accident. (It's not hard -- a year of extra life is a lot of hours.)
It ticks people off if they think that you're telling them what to do. So, instead, pretend to drive like you were drunk, or falling asleep. Let the car slowly drift to the side of the road, hit some rumbly stuff, and suddenly jerk it back to the center (this is my friend W's idea). Or, to be mathematical, use piecewise linear driving -- no smooth curves and adjustments, just a sequence of straight lines, punctuated by jerky corrections.
Up here in the northeast, you can help out the (non-existent, it seems) street sweepers, by driving very very close to the curb, and pick up the sand and gravel there, to get it out of the road. Driving through mud puddles works, too. Not too many people seem to like to drive behind that.
And once, in Texas, somewhat late at night on US290 between Austin and Houston, someone tailgated me in a pretty creepy way. I leaned down and reached under the passenger seat to retrieve the breaker bar I kept there, just in case -- didn't show it, either -- and the guy behind me backed off all of a sudden.
Add to that list, people overdriving their abilities in poor conditions. I drive, I bike. About 30 minutes BEFORE sunset (a full hour before Massachusetts requires that headlights be on) it gets dim enough that I can see my wimpy bicycle headlight painting reflectors. That's really dim; when building my wimpy lights, I spent a little time reading about human vision and color, and one thing that popped out is how terrible our vision is at dusk. Do people slow down proportionately? Not at all. Do they turn on their great big non-wimpy car headlights? Mostly, no. A half a square foot of high quality reflectors on the back of my bicycle, wasted.
It is also worth noting, that if your pool of transportation risk factors includes "heart attack", "stroke", and "diabetes" (these are slightly more common than exploding trucks), then the safest vehicle on the road (at least for the first 100 miles of each week's travel) is a bicycle. Modern cars, besides lulling you to sleep, lull you to fat. The estimates I have read (and I am still trying to track this down to the original data, but the description of the methodology sounds good) come from Mayer Hillman, comparing years of life lost by cyclists and non-cyclists, from diseases of the unfit, and from crashes. Bikes win by a factor of 10 to 20. If you haven't taken care of basic weekly exercise, but still worry about the finer points of high-speed auto safety, you're not allocating your concern/time/attention wisely.
Slashdot Mirror
True. But happily, there is data, and the data says that the expected lifespan in Canada is 4 years longer than it is in the US, the infant mortality rate (deaths per 1000) is 5.1 instead of 6.3, yet they spend $2000 less per capita, and instead of spending 15.4% of GDP on healthcare, they spend 9.8%.
There's your data -- the Canadian system is better. Your friend's dead uncle is just an anecdote. Perhaps the surgery he 'needed' was highly unlikely to extend his life.
Obviously we need more gay doctors.
Or women doctors, perhaps?
A bigger insurer (ie. Uncle Sam) won't fix it.
Most modern countries with universal health care get better outcomes (healthier, lower infant mortality, longer expected lifespan) for less cost per capita and a smaller percentage of their GDP. In their case, it appears that a bigger insurer worked pretty well, plus they don't have anyone going bankrupt from medical bills. If it works for Canada, England, France, Germany, Japan, Australia, Switzerland, Sweden, Israel, Iceland, New Zealand, Italy, Spain, Norway, Greece, Austria, Netherlands, Belgium, Finland, and South Korea, it can probably work for us, too.
(source: http://www.nationmaster.com/ who gets their data from the UN, OECD, CIA Fact Book, WHO, and similar reliable sources).
Cannabis is a good deal less toxic and/or addictive than quite a few over-the-counter drugs. Its side-effects (primarily, intoxication) are common, but not life-threatening. Against painkillers, it compares especially well.
I do agree, however, that if it were marketed as a drug, the guys selling it would find some way to push the truth envelope.
But this is not exactly what charging by the bit does. There are two problems.
#1, charging by the bit does not relieve the immediate congestion. Instead, it creates a customer-annoying surprise at the end.
#2, this is very, very time sensitive, and customers don't have a good way of knowing what is going on at which time. If the network load is under its maximum practical capacity, there's no problem to solve, and charging by the bit is merely a handy way of getting extra money. If the network is at capacity, there is no particular guarantee that the people who are clogging it, are your multi-terabyte P2P downloaders -- perhaps they confine their big slurps to the late evening hours, perhaps a bunch of otherwise light users all decided to see the live Victoria's Secret show. Charging the P2P users for their off-peak consumption, or killing all the P2P connections, doesn't do anything to reduce the bottleneck; in fact, the incentive is completely disconnected from the people causing the problem.
It would be nice, but is not yet practical, to have something like "auction-class" bandwidth. People get (say) $40 worth of micro-tokens to spend on bandwidth, and browsers, P2P, etc, bid for bandwidth. Configure the browser to know roughly what your cost preferences are (based on browsing history) and if you go to visit a movie that exceeds your usual cost, or if you start watching a lot of movies in peak hours, the browser can let you know that you're starting to spend too much "money". Once your monthly "money" is spent, you're stuck at bulk rate. And P2P guys, they would prefer bulk rate in general, so they would make "bulk rate" a pretty unpleasant place to be. This could have its own problems -- ISPs could dodge network neutrality by charging a higher rate to access Google, in hopes of forcing Google to pay some sort of ransom. But, lacking "auction-class", you can keep track of who is the source of whatever traffic is spiking capacity to the limit, and you can (if I understand TCP correctly) slow them down. This might well have the effect of blocking P2P during peak hours, but P2P works just fine off-peak, too.
There's two inconsistent lines of argument here. One is that it is all at the last mile (in which case, it becomes technology dependent. With DSL, as I understand it, the last mile is "my" piece of copper wire), and everything is cheap at the datacenter.
The other claim is that the problem occurs at the datacenter, when everybody wants to watch the bandwidth at the same time. So, which one is it?
My suspicion, not incredibly well informed, is that it has more to do with competing lines of business. After all the whole P2P throttling thing that Comcast did (which they say they are not doing now, though oddly enough, P2P remains broken for me) could have instead been invested in simple peak/non-peak bandwidth-based throttling (i.e., in a protocol-agnostic way, at any bottleneck in your network, penalize the guy(s) who are actually spiking the demand). They chose, instead, to do all this deep packet inspection nonsense. Why that, and not a more direct regulation of delivered bandwidth exactly when it was a problem?
They're coming pretty steadily. I read the FA, logged in to the ssh-receiving box, looked at the logs, and a half-dozen showed up while I was looking. Makes me wish I had the time to run a honeypot.
Based on my quick reading of the ssh/sshd documentation, I created an empty ~/.ssh/rc file for the accounts with known-good passwords (i.e., mine) and then put "kill -9 $PPID" in /etc/sshrc.
That seemed to work, though I know it will be bad for X11 (which I never use over ssh to my home box).
There may be more elegant ways to do this, but if there are, I didn't find them quickly (which is to say, the ssh documentation, like all Unix documentation, sucks rocks).
What I understand is, that LEDs will get cheaper in a "few" years. Apparently there's a little bit of expensive sapphire in each and every power LED, and replacing that is the key, and at least two different research groups have made progress on it. The next hurdle after that is building a good-looking and/or compatible fixture that keeps the LEDs cool enough. Overheat them, their output and their lifetime both go down.
There's also economies of scale that have yet to kick in, and there's the (as we have seen) charlatan factor at work. Enough people will sell crap LEDs to give the good ones a bad name.
The drive to conserve products costs also drives some pretty silly economies. I've been playing with bicycle lights (go look at some prices, be appalled), and if you are willing to pony up for a "generator" (really alternator) hub, you can, for under about $100 in parts, light your bike up like Las Vegas. You can run a standlight, you can charge batteries, you could run a little cell-phone charger. But if you look at what is offered for sale, it's diddly-shit. The "standlight" you get, if anything, is this itty-bitty thing run from a capacitor, not continued operation of your running lights at rated power. Companies are assembling the minimum, and selling it for what the high-end market will bear, which is surprisingly much.
I'll do the best I can here. Short answer is, LEDs are still too expensive for general use, but they already win for certain applications. The three places where I think they win are (1) under cabinets, (2) loading docks, and (3) bicycles.
The best LEDs are now a hair more efficient than fluorescent, but you won't want to use the best LEDs unless you enjoy reading by the light of an arc welder. "Neutral" LEDs are a hair less efficient than fluorescent, but the light is all sent to one plane (assuming you get one with a Lambertian lens on the LED, which is usual for power LEDs, e.g., CREE and Luxeon) which I think works to their advantage. In general, if you need N watts of fluorescent light, you need N, or nearly N, watts of LED light as a replacement.
Conversion power losses, which you mistakenly quote from the headline, are about the same. Cheap wall warts are about as bad as CFLs or cheap fluorescent ballasts.
LEDs are claimed to last 50,000-100,000 hours, and that is not burn-out, but reduced efficiency (70% of rated, I think). Better cooling increases lifetimes; they like to not be hot, and heat sinking is actually a major issue. This is sort of a pricing problem, because the temptation is to economize on the LEDs and up the current (the ones I like for indoor use are rated for 350-700mA, but you don't get double the light at double the current). At the 350mA loading, each 1W LED (bought retail, mounting on hexagonal pucks) costs about $8. 50 watts of LED, which would be a buttload of light, would be $400, plus drivers (you'd need 6 of the ones that I've used, and at $30 apiece, that's another $180). So when I say "expensive", I mean it. You can do better than my estimates, but it would require work, planning, design, etc.
To my eye, LEDs have a better light. There's at least one person here claiming otherwise with great authority, but I think they are working from faulty information or cheap-N-sleazy LEDs.
The reason that LEDs work well for undercabinet lighting is that you don't need to do anything special for them; they are so low profile, you can just tack them up bare. The light goes exactly where it is needed, and that sort of lighting is generally not so cheap so they are more nearly cost-competitive. For loading docks, LEDs are durable, and come on quickly in the cold, and don't need frequent replacement. On bicycles, LEDs are durable enough (CFL on a bouncing bike? I don't think so) and very efficient, and work at a low non-scary voltage.
Poke around at ledsupply.com, they have a good sampling of LED-related stuff.
Depends a whole lot on how cold your "cold" is, and which CFLs you use. We have a couple in ceiling fixtures, that in a Boston winter, get kinda cold, and start pretty slow. They're definitely less than five years old, because the ceiling is less than five years old. But I am pretty sure that these are also cheap bulbs, mostly because there's no assurance that an expensive bulb is actually any better, even though some of them might be. ("We promise our CFL doesn't suck" -- if I were a cheap CFL huckster, I'd slap that label on my crappy merchandise 15 seconds before I tripled the price.)
Exactly true. I checked the power factor on the wall wart running some LEDs with a kill-o-watt, and it was something like .6. The problem is entirely in the wall wart, of course, since the LEDs draw DC.
Most noticeable with Amber -- if you swing the junction temperature from 100 c to 0 C, the light output goes up by a factor of 4. (So says the Luxeon III data sheet (ds46), page 9).
I think your information is basically wrong. Perhaps you've been buying the wrong lights. The lumen ratings for bulbs take the human visual color response into account, so a "white" LED of 100 lumens is dumping a pretty good load of light into the green range. The color temperature on the most efficient LEDs is a hair off (looks like a welder's torch to me), but their light is as good as what you get from similarly efficient fluorescent bulbs. The LEDs I have in mind are the relatively recent CREE XRE power LEDs, that run at between 350 and 1000 mA.
If you are willing to take a minor efficiency hit, go for a quality "neutral white" LED, and you get pretty good light. I did this for the lights under our kitchen cabinets, and it is great.
You are correct that there are no efficient "direct" green LEDs, but the phosphor downconversion is no different from how it is done in fluorescent lights, and because the initial light is not confined to a small number of wavelengths, the resulting light is better looking. I initially got this information from a physicist (we were taking our kids on the same ski trip) and I later confirmed it for my own self with a diffraction grating. By-the-way, if you look at those diffraction-grating pictures, you will see that the white LEDs do indeed emit a good amount of yellow and green.
If you see flickering in your LEDs, you are using a truly cheap regulator. All the ones that I use are built around switching power supplies for efficiency, and their "flicker" is in the 10-100 kilohertz range. As far as durability goes, I have generator-driven current regulated LEDs on two of my kids' bikes and two of my bikes, and they ALL live outdoors, one not under any cover, in Massachusetts, year round. The kitchen lights are on almost continuously, since they are the most efficient lights in the house by far. Most important to long lifetime, is adequate heat sinking, and/or not over-driving the lights (they're also more efficient at lower power, even holding the temperature constant, and they're more efficient at lower temperature, which is a side-effect of lower power when you don't hold the temperature constant).
Where the heck did you get your information? If this is from lights that you've bought/used, I'd love to know which ones, so that I will never buy them by accident.
I don't quite understand your analysis -- the reasoning behind Hillman's conclusion is, as I understand it, that non-cyclists (as a group) tend to not get enough exercise, and the risks of THAT (insufficient exercise, as a group) are 10 times higher than the risks of getting killed on a bicycle. To point out that in fact, the crash-death risk of cycling is higher than the crash-death risk of caging does not refute this, unless we are somehow less dead when we die from the heart attack or stroke that is (apparently) much more likely to kill the cager. (And note that Hillman is counting expected years of life lost, not deaths.)
All that said, there are safer and less safe roads, and of course the prudent cyclist will take care to avoid obvious dangers. There's hills I won't descend down (I coast at 35+mph, I measured, and the road is skinny with poor sight lines), and there's roads I avoid in the dark (again, poor sight lines, poor visibility). But note, in all cases, that it is not he bicycle itself that is dangerous -- I avoid the hill because a car might not see me, I avoid the skinny curvy dark road because a car might not see me. The danger comes from the cars, and it would be lovely if more people recognized this.
I'm not sure where this Segway gadget fits in. It still seems to use plenty of energy (35mph, with that big fat profile, is going to use a heap of power, by bicycling standards), and it hardly has a monopoly on being powered by electricity (there's some practical, not-too-expensive, e-assist systems used in cargo biking). It looks heavy, and being wide like that, it will be prone to getting stuck in jams. If the batteries fail, your options are not so good. And it will encourage people to not get the exercise that they need, so that risk (which is apparently enormous) is not reduced.
Which metro area do you prefer, Tampa/St.Pete, Houston, Silicon Valley, or Boston? I ride about 50 miles per week, about 40 of that is on roads shared with cars. But it is true, Boston is well-known for its careful, considerate drivers, so perhaps things are different elsewhere.
The risk estimate is not a silly statement. The estimates are that driving to the exclusion of biking, is 10 to 20 times more dangerous (in terms of expected years of life lost) than biking. This was measured in England (the cycling risks vary from country to country, but so does obesity) by a guy named Mayer Hillman at the Policy Studies Institute at the University of Westminster. Biking in England is less safe than northern Europe, but safer than the United States.
I think you need to do a more rational assessment of risks. There are safe roads and unsafe roads, but overall, not getting enough exercise, is very, very unsafe, long term.
You could move to someplace civilized :-). And yes, I know, some places are just screwed up. Riding in Florida as a kid was a mixed bag; the heat makes you sweaty, but warm rain is no big deal. Timing helps; if there's an oil shortage (as there was then) the jerk fraction goes down nicely.
Walking's relatively dangerous, at least according to statistical studies. Makes no sense to me, but that's what the numbers say.
Bicycles can carry quite a lot of "stuff" -- mine (xtracycle cargo bike) is comfortable up to about 100 lbs, and it's rated for about 200. I do almost all shopping by bike nowadays. Others -- Long Johns, and Burrows 8 -- are more suited for serious lumpy loads, and put them nice and low.
You're taking your life in your hands when you drive everywhere, too, it just isn't so obvious. The risks from heart disease and other afflictions of the unfit are much larger that those from cycling, though perhaps your roads are unusually unsafe.
The weather problems of cycling are also overrated. The cars are much more dangerous in bad weather, so you have to be extra-careful of them, but clothing is a big help at dealing with the cold, and lack of clothing helps with the hot. Rain is, indeed, the worst.
Segways, on the other hand, what happens to them when you ride them through foot-deep water? How do their batteries cope if they're left out in 20-degree weather? Can I put snow tires on them? They look at little wide for threading through stopped traffic. And won't all that sitting around make me fat(ter)?
I must not be trying hard enough, or else the dispatch teams in places I have lived, are not as good as yours :-).
I both ride and drive, both on the road, usually. The health benefits of biking are largest in the first 50 miles per week, and with a 10-mile commute, kids, and weather, biking every day is impractical. When I drive, my first priority is not to hit people, my second priority is to not hit cars, my third priority is to obey the laws. Last on the list is get-there-fast. Even putting speed as the last priority, cars are plenty fast.
And if you don't think that you're competent to drive safely around bicycles (note that the car is the source of the danger, not the couriers), why are you driving? Seriously -- how can you say something as ridiculous as "or I will run you over"? Doesn't your car have a brake pedal? Don't you stop if things look uncertain, as if your car might hurt someone? You're wearing a 2000-pound suit of armor -- it's not like the courier can hurt you. Why aren't you being more careful that you don't hurt someone with your armor suit?
Your choice of domicile has no bearing on the relative safety of 85mph, 45mph, or 15mpg tailgating-induced crashes, so I don't quite see how that is relevant. If you were interested in maximizing your safety (including all risks), you would also get some physical activity several times per week. Speed of travel isn't everything.
And hitting a solid object at 85mph is also "often fatal", and you therefore argue that the risks are equal? Slashdot reasoning at its finest.
Oddly enough, 15mph is a respectable speed on a bicycle, and up to at least 50 miles per week, time spent on a bicycle is more than paid back in extended lifetime from improved fitness, even including the years laws to the rare bicycle accident. (It's not hard -- a year of extra life is a lot of hours.)
It ticks people off if they think that you're telling them what to do. So, instead, pretend to drive like you were drunk, or falling asleep. Let the car slowly drift to the side of the road, hit some rumbly stuff, and suddenly jerk it back to the center (this is my friend W's idea). Or, to be mathematical, use piecewise linear driving -- no smooth curves and adjustments, just a sequence of straight lines, punctuated by jerky corrections.
Up here in the northeast, you can help out the (non-existent, it seems) street sweepers, by driving very very close to the curb, and pick up the sand and gravel there, to get it out of the road. Driving through mud puddles works, too. Not too many people seem to like to drive behind that.
And once, in Texas, somewhat late at night on US290 between Austin and Houston, someone tailgated me in a pretty creepy way. I leaned down and reached under the passenger seat to retrieve the breaker bar I kept there, just in case -- didn't show it, either -- and the guy behind me backed off all of a sudden.
And the accident that results at 85mph, is no more lethal than the one the one that results at 15mph? Speed has no role in this at all?
Fuck yeah ditto.
Add to that list, people overdriving their abilities in poor conditions. I drive, I bike. About 30 minutes BEFORE sunset (a full hour before Massachusetts requires that headlights be on) it gets dim enough that I can see my wimpy bicycle headlight painting reflectors. That's really dim; when building my wimpy lights, I spent a little time reading about human vision and color, and one thing that popped out is how terrible our vision is at dusk. Do people slow down proportionately? Not at all. Do they turn on their great big non-wimpy car headlights? Mostly, no. A half a square foot of high quality reflectors on the back of my bicycle, wasted.
It is also worth noting, that if your pool of transportation risk factors includes "heart attack", "stroke", and "diabetes" (these are slightly more common than exploding trucks), then the safest vehicle on the road (at least for the first 100 miles of each week's travel) is a bicycle. Modern cars, besides lulling you to sleep, lull you to fat. The estimates I have read (and I am still trying to track this down to the original data, but the description of the methodology sounds good) come from Mayer Hillman, comparing years of life lost by cyclists and non-cyclists, from diseases of the unfit, and from crashes. Bikes win by a factor of 10 to 20. If you haven't taken care of basic weekly exercise, but still worry about the finer points of high-speed auto safety, you're not allocating your concern/time/attention wisely.