crosses a boundary
If there were no gravity, that would be correct if you meant "moves into a region with a different refractive index(speed of light)". If you form a gradient of RI, you can run the light around in smooth curves, instead of the instantaneous direction changes at boundaries. Most common example seen is cold ground with warmer air above it and no wind, big gradient forms. On a few mornings, I've seen Longs Peak (northern front-range Colorado, USA) get stretched up to where it looks half again as tall and steep. I thought it was an illusion, so I noted where the summit lined up with trees from a particular location (it was on my way to work), and checked it again from that same spot. When it's stretched, you can see features that are normally hidden by the foothills.
I'll also point out another non-straight travel for light - gravity lensing. I suppose you can say that it's going straight through the space it's crossing, but geometrically, it sure looks like a bend.
And it is 100% accurate
on
Crime Prediction
·
· Score: 2, Insightful
Well put, though incomplete.
It is rare that productivity actually benefits from a long wide-eyed zombie session. I mean... it does happen, when a large solution comes through and you are so excited to do it that you can't think of anything else anyway. Most of the time, you can keep working productively without your fingers on the keyboard and your eyes on the screen more than maybe 10% of the time.
Note: If your typing sucks, you'll get less rest time. One day, a friend came by my cube and looked over my shoulder. He apologied, explaining that he was annoyed by my "playing" with the keyboard, thinking that I was just drumming my fingers on the keys, but was surprised to see that it was all accurate text. He said I'd be dead silent for a couple of minutes, then maybe 20 seconds of a sort of "ripping" sound. Spend some time thinking between console sessions. I go maybe 10 minutes, then force myself to stand up. I either do a couple grass pickers, a pushup or two, or take a fastwalk lap around the office. Regardless, I do a few focus shifts - infinity to nosetip and back.
Here's an setup change for you, though... Get your monitor down as close to desk level as you can. It seems like everybody wants their monitors up high, but unless you're posing for an office products catalog, it forces you to open your eyes wide. That can be disastrous when you're "becoming one with" a screenful of code, and you forget that you're even made of meat. The less of your cornea that's exposed when you're an unblinking zombie, the closer exposed sections are to the wet edge of your eyelids, and capillary action has a fighting chance against evaporation.
Set up to have the monitor as far away as practicable. I gave up some density for range... 21" monitor, at only 1024x768. It's nice to get more info on the screen at one glance, but rarely necessary.
The best analogy I can give is to compare how you feel after crouching for an hour versus walking, standing, sitting, and laying down for 8 hours. Let your eyes move around, internally AND externally.
Oh, and don't plug your tear ducts. You make a place that stays warm and wet, without getting rinsed. Unless you're trying to grow bacteria in your tear ducts, that's not a good thing.
It doesn't have to simulate the entire universe as I assume it exists, only the part I can sense at any one time. All you imaginary people are paged out and suspended when I'm not aware of you. Besides - the simulation doesn't have to be "realtime". Time itself is part of the simulation. The results don't suffer just because it takes 2 hours to run me through a really complex second, or 8 hours of sleep completes in 11 femtoseconds.
The only case wherein simulation subjective time would need to match objective time would be if the simulation had to be part of objective reality - a training simulation, for instance. On the other hand, the storage requirements would be huge.
Want to live forever? Program a realtime simulation of yourself to start at the instant before you start it. That way, it will do the same, which will do the same. Your cpu time will be spread across an infinitely-growing number of simulations of yourself, causing all progress in your simulation to cease. You won't know it, though. I suppose you could add an interface to the simulation allowing it to know that it is a simulation, and increment a counter, limiting depth. You want to extend your lifespan 100 times? Run 99 simulations deep. 10^^6? 999999 subsims. You wouldn't notice the change, unless you caused a kernel panic, and probably not even then - rollback, in-memory edit, and you'd never know you'd even thought of it, nor would you ever again.
CAD? Perhaps Synopsys will do? They've got that, and their timing extraction tools, and others (I'm out of the semiconductor business now, so I've lost track), and last I knew, Cadence was porting their stuff too.
6 months of uptime with any modern Windows... means you're not installing the weekly critical security update patches, all of which require reboot. What's your IP address again?
I was expecting something along those lines, too. Back in the mid-late 1980s (long before the Tim Allen Show), I worked in a cabinet factory. Occasionally, I would line up a pair of Makita belt sanders... they had 50 foot cords, so I'd unplug them, pull the trigger and set the hold-on, line them up, and plug them in. The toughest part is getting a decent holeshot. It's a powerful, high-torque motor, and they're a bit top-heavy. In fact, if the belt isn't a little bit worn, they'll do a back-flip. It's not nearly as bad as you'd think, because the main mass of the motor is reversed from the belt rotation, so they're somewhat self-stabilizing, front-to-back. Getting the center of traction directly under the center of gravity is the killer, since the traction patch is somewhat dynamic, so what's good for the crucial first 6 inches is not so good for the next 3 feet, after which, the belt is usually running at ground speed. Most races are decided in the first foot anyway, as one (or both) vehicles usually spin and roll over on launch.
Note: While these things ARE toys, they are dangerous ones. A quarter-second grind on flesh leaves a big divot, and a finger sucked into the back roller isn't coming back.
Instead of superglue, clean the head... In such circumstances, I just use 70% isopropanol. It's a good solvent for hydrophobic and hydrophilic contaminants, and any leftover dries really clean. Spritz an appropriate-sized philips-head screwdriver head with silicon lubricant. Mix up a little JB-Weld, and work a drop into the head with a toothpick to ensure that it is contacting as much as possible. Set another drop on top of that one and stick your slicked-up screwdriver bit into it - secure with something... I like the little radioshack extra-hands jig thingy. GO AWAY, and leave it alone for a whole day. Then, turn the screwdriver bit. The head WILL turn with it. I can't vouch for the shank. If the shank stays, you'll have to drill it out, or perhaps just do without that mounting point? If you drill it out, be very careful to remove only enough to get a tap in there to clean the threads... you'll probably have to go up one size, as you're unlikely to get it perfectly down the center. Keep a vacuum on the drillpoint to avoid deadly metal slivers.
On larger screws/bolts/studs, I like to get some heavy-walled steel pipe with an I.D. just slightly smaller than the head. Set that on the head, and weld a bead around the inside joint. Take a big old 36-inch Rigid pipe wrench, put a pipe on it if necessary, and move it. I did that once on the upper balljoint on a Dodge dart. It actually took the threads right out of the upper control arm - smooth as if it were machined. Phyrric victory, but I won.
The slowest Seagate 20GB drive i found consumes 24 watts. You'll still need an interface - cpu, memory, network...
You're talking about NAS. So far, as you note, that's always been an office thing. They're made with ready power and heat dissipation taken for granted. I can't imagine a company with a real need for such a specialized requirement (as in, you're about the only case it's ever been needed) needing "ask/." help. If it's a power availability problem - spacecraft maybe?... you're not looking for a COTS solution... there would be none. If it's a power DENSITY problem... consolidate all the drive needs in a server away from the environment you're measuring.
Incidentally, just so you don't embarrass yourself in front of your boss... wireless, at least in the most-commonly used sense, uses both electric and magnetic fields, most commonly at 2.4 or 5 GHz. Hell, conventional ethernet is something like DC-UHF, but mostly contained in the cable... perhaps something optical might be better-suited to your needs?
Frankly, I think I'd take the DC outputs of the sensors and route them as the center conductors of good-quality coax, and bring them all to a bank of A/D converters somewhere away from the environment being measured.... can you give some kind of detail about the environment?
Anyway, when you roll your own, you'll have to use something like a Hitachi microdrive. The 1GB drive consumes 8-1/4 watts on write. Maybe a transmeta-based motherboard won't break the rest of the budget.
I'd be glad to work on it. I need a job.
The thing I don't get is how this is claimed to be the first time this is done. Years ago, on a biology class field trip in southern Indiana (USA), a teacher pointed out a plant that took up salt from the soil, and said that the pioneers used to use it as a seasoning, and would sometimes toast (open burning would leave NaO instead of NaCl) it to ashes and leach the salt out of the ash for winter use.
That said, it's still pretty cool that there's a plant that grows in heavily saline soil and can be used to desalinate it. Maybe they can get a cold-climate variety developed and reclaim the marshes around Hudson Bay?
Agreed on "Type B". If someone with a camera had snuck up on my truck the night I lost my virginity, (and been really quick) I could have been the subject of "child porn" in some jurisdictions.
I don't think a reasonable person would consider 16-and-up "children" in this context.
I'll tell you how they'd "limit the damage". They'd eliminate the evidence, then start recording "deficiencies in your job performance".
Hit him first. Hit him hard. When the dust clears, and the rightness of your actions is a matter of legal record, you'll be much safer.
That's not to say that you're not still fscked. Nobody wants to ever get into that situation, but once you're in it, pick your best chance for surviving it.
I just assumed that since he let it go that it was just plain porn, not kiddie porn. If it was kiddie porn, and he didn't quietly and promptly bring the police into it before informing anyone else in the company, he is a child molester himself - not molesting with his own body, but assisting in the act.
One caveat: Some of the perverts stick that crap on legitimate porn newsgroups and such, with headers that do not indicate what it is. If there're only a very low proportion of kiddie porn among lots of legitimate porn, he may have just been hit with the equivalent of a goatse.cx troll. run diskreet or something like that on it, and check him again in a few weeks. If it's plain that he's seeking kiddie porn, put him UNDER the jail.
While I agree that it's probably time to drop that requirement, I do think there should be a seperate code endorsement giving access to narrow slices (that's all it needs) of each band, kind of like archery and muzzleloader hunting seasons.
My code sucks, but it still always gives me a thrill to get through clearly when SSB is just a wispy modulation on the noise.
anyone who can correclty identify this quote
anyone who can't is probably not reading/.
Anyway, keep your 5 points.
Re:A message from Capt. Picard:
on
Dancing Barefoot
·
· Score: 1
Gee... Do you suppose he doesn't hear that every day or so?
My wife and I both thought it was very funny on the show, that first time. Like everyone, we found the Wussley character to be an irritant, and the line was an obvious acknowledgement of fan annoyance.
He did a good job with an annoying character, and takes the ribbing like a man. It's been what, 10 years now? Perhaps we could let it drop? Acting is still his career, near as I can tell. It would be nice if he could get away from Wussley so he can work again.
It's a light bulb company that bought, and is gradually destroying, some cool electronics companies.
Their entire semiconductors division is imploding. In a cyclical industry, this cash-rich company has decreed that the division will not post a loss for any quarter. Their european-style management creates a social climate wherein all but frontline management personnel are immune to cuts. As a result, they are running fully 1/3 management at last count - that's right, 1 management person for every two producers. To meet their salary cost cut goals, they are now laying off fully booked engineers.
I had already seen incomprehensible decisions, which they are now being forced to compensate for at great expense and weak return, but cutting the profit centers to prevent loss says it all. I can infer only that they are intentionally disposing of their highest-tech division.
Mini maglights use a tiny bulb with straight parallel leads stuck in a pair of holes... just about the configuration of a standard LED. Buy a minimaglight, a 2.9something volt LED, and a pair of NiMH AA cells. I suggest the NiMH more for the lower voltage than for rechargeability. That lets you skip trying to fit in a resistor, which I've never managed. LEDs leads tend to be a hair skinnier than the ones on the lamps, so you will probably have to knurl them. I just lightly "bite" them, with vise-grips.
Pull the lamp out just enough to expose a bit of lead without disconnecting it, and use your VOM to see the polarity. Remove the bulb and gouge a little pit at the positive side.
Cut the LED leads to match the bulb.
Test fit. If it doesn't come on, knurl the leads and try again.
You'll probably have to ream out the reflector a little, as the LED is probably slightly bigger around than the bulb.
Mine's lasted for 3 years now, and goes about 20 hours on a charge.
I've seen little LED hiking lights recently, at Eastern Mountain Sports, for not much more than what I've got in the homebrew, though. They're smaller and brighter than mine, and they come with an elastic headband, which is awfully convenient. Most of them, though, use non-standard batteries, so the cost may get out of hand. I just don't know about that part.
I just figured I'd provide something a little kinder, though really less helpful, than what everybody else is probably suggesting.
I'd never heard of it until this post, but the context indicates "F(ree) D(ocumentation) L(icense)", which google confirms. Some acronyms are so obvious that I can forgive the editors for assuming their audience either already knows, or can look them up themselves. If a newspaper used an obscure term, name, or acronym without defining it, that's a real booboo, as the reader is unable to comprehend the article without a major context change. However, this isn't a dead tree document. If you're reading it, you also have instant access to other references.
That said, I too am troubled by the dupes. I'm perfectly capable of recognizing them, and combining the useful comments from each, but I'd expect the editors to regularly skim the content they are managing. It doesn't mean they're not doing a good job, but such slap-in-the-face obvious dumb mistakes as dupes and misspellings keep bringing up questions. It's like a big tough marine with a lisp. There's that constant cognitive dissonance between the fact that he is tough and the other fact that he sounds like a sissy.
The editors are all smart and work hard. CT is making money on the web without pr0n, for heaven's sake!...But, they keep doing things that make them sound stupid and lazy(in the negative sense).
Anyway, cut them some slack for overestimating you.
Slashdot Mirror
crosses a boundary
If there were no gravity, that would be correct if you meant "moves into a region with a different refractive index(speed of light)". If you form a gradient of RI, you can run the light around in smooth curves, instead of the instantaneous direction changes at boundaries. Most common example seen is cold ground with warmer air above it and no wind, big gradient forms. On a few mornings, I've seen Longs Peak (northern front-range Colorado, USA) get stretched up to where it looks half again as tall and steep. I thought it was an illusion, so I noted where the summit lined up with trees from a particular location (it was on my way to work), and checked it again from that same spot. When it's stretched, you can see features that are normally hidden by the foothills.
I'll also point out another non-straight travel for light - gravity lensing. I suppose you can say that it's going straight through the space it's crossing, but geometrically, it sure looks like a bend.
Isn't that right, Mr. Anderton?
Well put, though incomplete.
It is rare that productivity actually benefits from a long wide-eyed zombie session. I mean... it does happen, when a large solution comes through and you are so excited to do it that you can't think of anything else anyway. Most of the time, you can keep working productively without your fingers on the keyboard and your eyes on the screen more than maybe 10% of the time.
Note: If your typing sucks, you'll get less rest time. One day, a friend came by my cube and looked over my shoulder. He apologied, explaining that he was annoyed by my "playing" with the keyboard, thinking that I was just drumming my fingers on the keys, but was surprised to see that it was all accurate text. He said I'd be dead silent for a couple of minutes, then maybe 20 seconds of a sort of "ripping" sound. Spend some time thinking between console sessions. I go maybe 10 minutes, then force myself to stand up. I either do a couple grass pickers, a pushup or two, or take a fastwalk lap around the office. Regardless, I do a few focus shifts - infinity to nosetip and back.
Here's an setup change for you, though... Get your monitor down as close to desk level as you can. It seems like everybody wants their monitors up high, but unless you're posing for an office products catalog, it forces you to open your eyes wide. That can be disastrous when you're "becoming one with" a screenful of code, and you forget that you're even made of meat. The less of your cornea that's exposed when you're an unblinking zombie, the closer exposed sections are to the wet edge of your eyelids, and capillary action has a fighting chance against evaporation.
Set up to have the monitor as far away as practicable. I gave up some density for range... 21" monitor, at only 1024x768. It's nice to get more info on the screen at one glance, but rarely necessary.
The best analogy I can give is to compare how you feel after crouching for an hour versus walking, standing, sitting, and laying down for 8 hours. Let your eyes move around, internally AND externally.
Oh, and don't plug your tear ducts. You make a place that stays warm and wet, without getting rinsed. Unless you're trying to grow bacteria in your tear ducts, that's not a good thing.
It doesn't have to simulate the entire universe as I assume it exists, only the part I can sense at any one time. All you imaginary people are paged out and suspended when I'm not aware of you. Besides - the simulation doesn't have to be "realtime". Time itself is part of the simulation. The results don't suffer just because it takes 2 hours to run me through a really complex second, or 8 hours of sleep completes in 11 femtoseconds.
The only case wherein simulation subjective time would need to match objective time would be if the simulation had to be part of objective reality - a training simulation, for instance. On the other hand, the storage requirements would be huge.
Want to live forever? Program a realtime simulation of yourself to start at the instant before you start it. That way, it will do the same, which will do the same. Your cpu time will be spread across an infinitely-growing number of simulations of yourself, causing all progress in your simulation to cease. You won't know it, though. I suppose you could add an interface to the simulation allowing it to know that it is a simulation, and increment a counter, limiting depth. You want to extend your lifespan 100 times? Run 99 simulations deep. 10^^6? 999999 subsims. You wouldn't notice the change, unless you caused a kernel panic, and probably not even then - rollback, in-memory edit, and you'd never know you'd even thought of it, nor would you ever again.
That's "Mr. Plow", that's the name. That name again is "Mr. Plow"/B>".
Nobody reads the good stuff any more? This scientist - he be scowler?
Finally, it falls to me to say so.
CAD? Perhaps Synopsys will do?
They've got that, and their timing extraction tools, and others (I'm out of the semiconductor business now, so I've lost track), and last I knew, Cadence was porting their stuff too.
You actually know something about that show? You're so gay.
6 months of uptime with any modern Windows... means you're not installing the weekly critical security update patches, all of which require reboot.
What's your IP address again?
Coronavirus is from Mars, Chlamydia is from Venus.
I was expecting something along those lines, too. Back in the mid-late 1980s (long before the Tim Allen Show), I worked in a cabinet factory. Occasionally, I would line up a pair of Makita belt sanders... they had 50 foot cords, so I'd unplug them, pull the trigger and set the hold-on, line them up, and plug them in. The toughest part is getting a decent holeshot. It's a powerful, high-torque motor, and they're a bit top-heavy. In fact, if the belt isn't a little bit worn, they'll do a back-flip. It's not nearly as bad as you'd think, because the main mass of the motor is reversed from the belt rotation, so they're somewhat self-stabilizing, front-to-back. Getting the center of traction directly under the center of gravity is the killer, since the traction patch is somewhat dynamic, so what's good for the crucial first 6 inches is not so good for the next 3 feet, after which, the belt is usually running at ground speed. Most races are decided in the first foot anyway, as one (or both) vehicles usually spin and roll over on launch.
Note: While these things ARE toys, they are dangerous ones. A quarter-second grind on flesh leaves a big divot, and a finger sucked into the back roller isn't coming back.
Instead of superglue, clean the head... In such circumstances, I just use 70% isopropanol. It's a good solvent for hydrophobic and hydrophilic contaminants, and any leftover dries really clean. Spritz an appropriate-sized philips-head screwdriver head with silicon lubricant. Mix up a little JB-Weld, and work a drop into the head with a toothpick to ensure that it is contacting as much as possible. Set another drop on top of that one and stick your slicked-up screwdriver bit into it - secure with something... I like the little radioshack extra-hands jig thingy. GO AWAY, and leave it alone for a whole day. Then, turn the screwdriver bit. The head WILL turn with it. I can't vouch for the shank. If the shank stays, you'll have to drill it out, or perhaps just do without that mounting point? If you drill it out, be very careful to remove only enough to get a tap in there to clean the threads... you'll probably have to go up one size, as you're unlikely to get it perfectly down the center. Keep a vacuum on the drillpoint to avoid deadly metal slivers.
On larger screws/bolts/studs, I like to get some heavy-walled steel pipe with an I.D. just slightly smaller than the head. Set that on the head, and weld a bead around the inside joint. Take a big old 36-inch Rigid pipe wrench, put a pipe on it if necessary, and move it. I did that once on the upper balljoint on a Dodge dart. It actually took the threads right out of the upper control arm - smooth as if it were machined. Phyrric victory, but I won.
The slowest Seagate 20GB drive i found consumes 24 watts. You'll still need an interface - cpu, memory, network... /." help. If it's a power availability problem - spacecraft maybe?... you're not looking for a COTS solution... there would be none. If it's a power DENSITY problem... consolidate all the drive needs in a server away from the environment you're measuring.
You're talking about NAS. So far, as you note, that's always been an office thing. They're made with ready power and heat dissipation taken for granted. I can't imagine a company with a real need for such a specialized requirement (as in, you're about the only case it's ever been needed) needing "ask
Incidentally, just so you don't embarrass yourself in front of your boss... wireless, at least in the most-commonly used sense, uses both electric and magnetic fields, most commonly at 2.4 or 5 GHz. Hell, conventional ethernet is something like DC-UHF, but mostly contained in the cable... perhaps something optical might be better-suited to your needs?
Frankly, I think I'd take the DC outputs of the sensors and route them as the center conductors of good-quality coax, and bring them all to a bank of A/D converters somewhere away from the environment being measured.... can you give some kind of detail about the environment?
Anyway, when you roll your own, you'll have to use something like a Hitachi microdrive. The 1GB drive consumes 8-1/4 watts on write. Maybe a transmeta-based motherboard won't break the rest of the budget.
I'd be glad to work on it. I need a job.
The thing I don't get is how this is claimed to be the first time this is done. Years ago, on a biology class field trip in southern Indiana (USA), a teacher pointed out a plant that took up salt from the soil, and said that the pioneers used to use it as a seasoning, and would sometimes toast (open burning would leave NaO instead of NaCl) it to ashes and leach the salt out of the ash for winter use.
That said, it's still pretty cool that there's a plant that grows in heavily saline soil and can be used to desalinate it. Maybe they can get a cold-climate variety developed and reclaim the marshes around Hudson Bay?
Agreed on "Type B". If someone with a camera had snuck up on my truck the night I lost my virginity, (and been really quick) I could have been the subject of "child porn" in some jurisdictions.
I don't think a reasonable person would consider 16-and-up "children" in this context.
Goodness. You've put a lot of effort into justifying your predilections, haven't you?
I'll tell you how they'd "limit the damage". They'd eliminate the evidence, then start recording "deficiencies in your job performance".
Hit him first. Hit him hard. When the dust clears, and the rightness of your actions is a matter of legal record, you'll be much safer.
That's not to say that you're not still fscked. Nobody wants to ever get into that situation, but once you're in it, pick your best chance for surviving it.
I just assumed that since he let it go that it was just plain porn, not kiddie porn. If it was kiddie porn, and he didn't quietly and promptly bring the police into it before informing anyone else in the company, he is a child molester himself - not molesting with his own body, but assisting in the act.
One caveat: Some of the perverts stick that crap on legitimate porn newsgroups and such, with headers that do not indicate what it is. If there're only a very low proportion of kiddie porn among lots of legitimate porn, he may have just been hit with the equivalent of a goatse.cx troll. run diskreet or something like that on it, and check him again in a few weeks. If it's plain that he's seeking kiddie porn, put him UNDER the jail.
While I agree that it's probably time to drop that requirement, I do think there should be a seperate code endorsement giving access to narrow slices (that's all it needs) of each band, kind of like archery and muzzleloader hunting seasons.
My code sucks, but it still always gives me a thrill to get through clearly when SSB is just a wispy modulation on the noise.
anyone who can correclty identify this quote /.
anyone who can't is probably not reading
Anyway, keep your 5 points.
Gee... Do you suppose he doesn't hear that every day or so?
My wife and I both thought it was very funny on the show, that first time. Like everyone, we found the Wussley character to be an irritant, and the line was an obvious acknowledgement of fan annoyance.
He did a good job with an annoying character, and takes the ribbing like a man. It's been what, 10 years now? Perhaps we could let it drop? Acting is still his career, near as I can tell. It would be nice if he could get away from Wussley so he can work again.
-
It's "Philips", not "Phillips", unless you mean Phillips Petroleum
-
It's a light bulb company that bought, and is gradually destroying, some cool electronics companies.
Their entire semiconductors division is imploding. In a cyclical industry, this cash-rich company has decreed that the division will not post a loss for any quarter. Their european-style management creates a social climate wherein all but frontline management personnel are immune to cuts. As a result, they are running fully 1/3 management at last count - that's right, 1 management person for every two producers. To meet their salary cost cut goals, they are now laying off fully booked engineers.I had already seen incomprehensible decisions, which they are now being forced to compensate for at great expense and weak return, but cutting the profit centers to prevent loss says it all. I can infer only that they are intentionally disposing of their highest-tech division.
Mini maglights use a tiny bulb with straight parallel leads stuck in a pair of holes... just about the configuration of a standard LED. Buy a minimaglight, a 2.9something volt LED, and a pair of NiMH AA cells. I suggest the NiMH more for the lower voltage than for rechargeability. That lets you skip trying to fit in a resistor, which I've never managed. LEDs leads tend to be a hair skinnier than the ones on the lamps, so you will probably have to knurl them. I just lightly "bite" them, with vise-grips.
Pull the lamp out just enough to expose a bit of lead without disconnecting it, and use your VOM to see the polarity. Remove the bulb and gouge a little pit at the positive side.
Cut the LED leads to match the bulb.
Test fit. If it doesn't come on, knurl the leads and try again.
You'll probably have to ream out the reflector a little, as the LED is probably slightly bigger around than the bulb.
Mine's lasted for 3 years now, and goes about 20 hours on a charge.
I've seen little LED hiking lights recently, at Eastern Mountain Sports, for not much more than what I've got in the homebrew, though. They're smaller and brighter than mine, and they come with an elastic headband, which is awfully convenient. Most of them, though, use non-standard batteries, so the cost may get out of hand. I just don't know about that part.
I just figured I'd provide something a little kinder, though really less helpful, than what everybody else is probably suggesting.
I'd never heard of it until this post, but the context indicates "F(ree) D(ocumentation) L(icense)", which google confirms. Some acronyms are so obvious that I can forgive the editors for assuming their audience either already knows, or can look them up themselves. If a newspaper used an obscure term, name, or acronym without defining it, that's a real booboo, as the reader is unable to comprehend the article without a major context change. However, this isn't a dead tree document. If you're reading it, you also have instant access to other references. That said, I too am troubled by the dupes. I'm perfectly capable of recognizing them, and combining the useful comments from each, but I'd expect the editors to regularly skim the content they are managing. It doesn't mean they're not doing a good job, but such slap-in-the-face obvious dumb mistakes as dupes and misspellings keep bringing up questions. It's like a big tough marine with a lisp. There's that constant cognitive dissonance between the fact that he is tough and the other fact that he sounds like a sissy. The editors are all smart and work hard. CT is making money on the web without pr0n, for heaven's sake! ...But, they keep doing things that make them sound stupid and lazy(in the negative sense).
Anyway, cut them some slack for overestimating you.