Re:Shuttlecom have a fascination with Javascript
on
Shuttle SS51 Reviewed
·
· Score: 2
You must be looking very hard for the things you seem to want...
RadioShack springs to mind as a place to pick up an optical switch for toslink, fairly cheaply. Whether it uses mirrors, or close-coupled light pipes, or whatever, it probably works justfine.
Midiman sells a converter (called "CO2", IIRC) which converts between optical and coaxial SP/DIF. They pop up on Ebay from time to time, and work well for a variety of tasks (SCMS stripping, anyone?).
The inexpensive Audio Alchemy DDE v1.1 that I use as a DAC for my computer has coaxial and optical inputs, as well as an isolated, re-clocked coax output which just mirrors whatever input is selected.
Edirol has a product here which has five connections: USB, toslink in/out, and coaxial in/out.
Finally, as configurable as the sound support is on the Shuttle board, I'd be very surprised if one couldn't turn one of its three multi-purpose jacks into a coax output using nothing but a software switch.
Back In The Day, I made an informal study of this topic, on IRC (of course).
I chatted with prospective participants, gathered as much information from them about their modem as possible, asked if it were OK if they might drop carrier, and sent three magical ICMPs their way, one second apart. I logged the data in a text file, which I don't seem to be able to find just now.
Something like 80% were vulnerable. I was amazed. I gave my SupraFAXModem 288i a big fat kiss for being such a good sport about it all.
Hacks like this have existed for a -long- time. I remember helping sysops in 1991 to defend against kids who would dial up BBSs, and use the login prompt's echo to drop the board's modem to command mode, whereupon they'd leave it in a strange state and things wouldn't work until the sysop reset the modem.
Fun stuff. It wasn't until some years later that I learned about the Hayes patent on having a silence period between +++ and command mode.
As some other posters have pointed out, you don't need to have BIOS-level access to the entire hard drive to get the machine to boot properly and run Linux. Just keep the bootable partition within the first 500 megs, and all is dandy.
However, you might want to use a BIOS overlay, anyway. Makes things easier, and lets you do stuff like run ancient DOS games on the hardware they were meant for, or fire up OS/2 Warp for an old-school look at the future.
On my 386SL-25 laptop, I'm using IBM's overlay software, which is freely downloadable as a bootable floppy disk image. Most other manufacturers also supply overlay software, free, but it generally requires you to install it on a machine with a drive of the same make; Maxtor's software needs to see a Maxtor drive somewhere in the system, or it will simply refuse to cooperate.
This works fine with Linux, and has for a very long time. It just recognizes that the drive was partitioned with overlay software, and does the same sector translation on its own.
I doubt there's even a speed hit.
I used IBM's software because I have a bunch of their SCSI disks in the machine I was borrowing to do the Linux install with, and Hitachi stands almost alone as a vendor who doesn't supply overlay software of their own. (the laptop, sadly, has no floppy drive, CD-ROM, or other external storage, so I spent most of an evening swapping IDE cables trying to get the thing to boot.)
The component most likely to complain about temperature and vibration is the hard drive.
While your car's suspension will work rather well at isolating things from bumps in the road, it wouldn't hurt to get a little creative with mounting the drive. Suspending it between rubber straps might be a good idea, but watch out for things which are bouncy (resonant) -- it almost doesn't matter how much of the initial impact that your suspension absorbs, if it continues to shake the hell out of the thing several seconds after stuff should've come to a rest.
Hard drives come packed in open-cell foam, like a mattress pad, and seem to survive UPS Ground pretty well in such an arrangement. It wouldn't take too much creative engineering to fabricate an enclosure made predominately out of foam, but with a fan and enough room for air to circulate.
Temperature doesn't look like it's too much of a hassle, these days. I'm looking at the specs on an IBM 120GXP, which show it to be happy from -40 to 55 degrees celsius (-40 to 131 F). Since these are ambient temperatures, the drive is thus designed to withstand a trunk at 131 degrees with moderate airflow.
Since you'll be opening your windows and/or turning on the air conditioning Right Away on such blazing hot days, the hard drive should start recieving cooler ambient air at about the same time it starts generating appreciable heat of its own.
And in the winter, at -40, the last thing you're worried about is whether your hard drive will spin up. More important is whether or not the engine will, and if you'll be able to get the tires un-frozen from the ground. If you've got half a brain about you, you'll have the computer and stereo off, anyway, until the car is well under its own power...and, by that point, generating heat to warm up the electronics.
I'd avoid connecting the enclosure directly to the car's ductwork. Ever see a windshield fog up on the inside on a cold morning, just after you turn on the defroster? Imagine that happening to your in-car PC. Bad news.:-/
Since the motherboard you've chosen is so bloody small, have you considered putting it under the front seat? Things would get impossibly tight in a sports car with power seats, but should be do-able in almost anything else. There's no reason for the box to be more than an inch or three high. Sescom is a company who makes a large variety of metal boxes for do-it-yourself projects - chances are, one of them would fit your motherboard, hard drive, and some manner of DC power supply justabout perfectly, while remaining small enough to slide under the seat.
Drill or punch holes in the sides of the box, and mount the biggest, lowest-RPM fan you can find on the top of the box, blowing down. You might even be able to do away with having a dedicated CPU fan, and get by with a just large heatsink.
There's a few other things you might want to look into, with software. There's a way to poll hard drive temperature using a protocol called SMART - if the drive is cold, keep it spun up to reduce the viscosity of the grease in its bearings. Likewise, if the drive is fairly warm, spin it down when not needed to reduce wear and help ward off data loss (they are, of course, -much- more durable with not spinning).
Do the same with the CPU, if it suits you. If it's cold out, run the CPU full-tilt (seti@home, some random busyloop, cat/dev/zero >/dev/null, or whatever) to keep the motherboard components warm and well within design spec, which will help reduce condensation. Do the opposite on hot days - HLT the CPU when idle, so as not to make things any warmer than needed to get the job done, and take whatever other power-saving measures you can.
The goal here is to bring the board to some desired operating temperature, and keep it there until the car turns off.
All that said, you'll probably find that the most sensitive component of your system is the LCD display, which will be painfully slow/frozen on cold mornings, and either solid white or black after a hot day of sitting in the sun...and there's really not much to be done about it, unfortunately.
A 64x CAV CD-ROM drive will spin at a maximum of 12,800 RPM, according to my math, but drives don't seem to exist at that speed just yet. More common 56x drives spin at 11,200 RPM.
If things are consistantly and violently exploding at a little over twice that speed, would it not make sense that there is a very real safety issue in making things faster than this?
One might theorize that off-balance discs, cracked/scored/otherwise-damaged media, and just plain bad luck might cause things to go dangerously amiss even at current speeds.
Current high-end SCSI hard drives spin at 15,000RPM, but do so using extremely well-balanced, carefully-produced, expensive solid aluminum platters and motors. And, besides, they're also encased in heavy metal boxes, and don't have a soft plastic face through which to fire shrapnel into the chest of the user.
Consider that a CD-ROM has a much larger diameter than a typical hard disk platter, and is thus exposed to far greater centrifugal force and linear velocity. Consider also that a CD-ROM drive only costs a few dollars to make, and that CDs are down to a couple of cents each in large volume.
Given this information and that contained in the article, I doubt it would take much effort to make a CD explode in a current 56x drive, thus presenting a very real bottleneck, indeed.
Plextor drives can be 'underclocked' with software switches fairly trivially, using free tools. They also let you do some fanciness like control spin-down times, and how the drive responds to errors.
OTOH, the fastest drive Plextor makes operates at 40x CAV. And I've fed many, many messed up (cracked, deeply scratched, off-balance) CDs into my 32x Plextor, without ever having one disintegrate, even when they're spinning at high speed for 12 hours or more.
So, personally, I'm not too worried about the safety aspects. But if you want to slow down a CD-ROM for whatever reason, Plextor is a very sure route to follow.
Scary prospects, indeed, but not much (if any) worse than the current situation. There's a number of mechanical components in conventional vehicles with no redundancy and a tendancy to fail now-and-then.
If a tie rod snaps on a 1985 Buick LeSabre, you're quite simply fucked, unless you happen to be driving slow, in light traffic, or sitting at a stop sign or something.
If the rear hub assembly (think "spindle" or "wheel bearing") snaps in two on a 1995 Chevy Beretta, you're also probably fucked. The body bottoms out on pavement, the car turns around in a quick 270, and tends to disarm any attempt at serious braking. Half of one side of the rear brakes end up scattered along the roadway, and the other half stay bolted to the wheel as it rolls away and bounces off of a guard rail. This completely disables to the mechanical brake, and makes the hydraulic brake largely ineffective as fluid pours out of the smashed brake piston -- the pedal goes to the floor with remarkable ease.
Or, take ball joints as another example. Wheel ends up flopping freely around, and the strut now entrusted with keeping it attached gets ripped apart. Steering is iffy because one of the front wheels is locked hard right and smoking, power is gone (the half shaft gets ripped out of the transmission), and there's a good chance that a brake line will be snapped, making stopping rather interesting. (1993 Mercury Villager.)
And, of course, tires go flat. They explode, they remove themselves from the rim, they delaminate - and those not mounted on Ford Explorers are not exempt from this. I tore one off of the rim on the aforementioned Beretta while going sideways through a ditch, into a field of corn stubble at 80MPH (yep, that one was my fault).
This stuff happens, it's usually fairly spectacular, very fast, and there's no redundant parts in an automotive suspension or steering system to ensure continued control of the forward progression of the vehicle when these components fail.
Interestingly, the one fly-by-wire mechanism in that Chevy (the fuel injection system) has never shown any signs of ill intent.
Given that last point, and the failure rate of mechanical parts, I have no problem entrusting my life to a simpler system of electronic devices. That car has been through numerous water pumps, a couple of alternators, and various suspension parts (some more than once), but I've never had issue with any electrical system it has.
It is also worth noting that the power brakes and steering you blame for your inherent mistrust are also strictly mechanical systems...
I have no numbers to back this up, but in my experience, 2.4 -feels- slower on older hardware than 2.0.
'Sides, unless it's a router box and you need the latest, greatest QoS tools and security fixes on your 486, there's very little reason to upgrade.
Most people seem to get new kernels when they need support for new hardware, and that's just not much of an issue for a 486.;)
I did upgrade the kernel on my 386SL/25 laptop recently, from 2.0.37 to 2.2.18, but only because I wanted to play with the swsusp patches and didn't feel like learning how to backport them to 2.0.
The temptation to move to a 2.4-AC kernel with swsusp built-in was very easy to resist.
Ah, but you forget that SCSI is backward-compatible. Which is to say that every SCSI card is a SCSI-1 card. Even cards supporting Ultra320, capable of transferring 320 megabytes per second, using balanced signalling at 160MHz on a bus 16 bits wide will happily and efficiently step down to the 5MHz, 8-bit single-ended bus most people call SCSI-1, as long as things stay terminated properly.
This is also to say that a 15KRPM Ultra320 SCSI disk will happily work on a 15-year-old Mac, or an Amiga or Atari or IBM PC(tm) with some manner of SCSI interface, given only the correct adapters and termination.
In instances where fsck begins whining about things being seriously wrong, I politely tell it to fuck off. The machine then boots up with the partition in question mounted read-only.
This gives me a shot at doing a quick backup of any Really Important Stuff to tape, CD, or just another drive.
This hurts nothing in the process, but does cost a bit of time.
After that's done, I'll tell fsck to go ahead and smash the drive. It usually works great, at least on Linux, but now and then has issues which would've trashed my Really Important Stuff.
No big thing, though. Restore the Really Important Stuff from the freshly-made backup, and rebuild the rest of the filesystem by hand.
In a properly-implemented system of this type, each listener would provide a stream for n additional listeners, only when actually listening (where n is a user selected integer of fairly small value). If n=>~3, one would be able to help out the fraction of listeners who are hopelessly and forever firewalled and thus unable to mirror the stream themselves.
At any rate, you turn off the "radio" and go to bed - and all bandwidth consumption halts until such a time as you feel like turning it on again.
You shouldn't look at it as paying for someone else's music - that someone may very well be in the same boat as you, and must pay seperately to -recieve- the bandwidth you give them. Not to mention the poor soul who was kind enough to let you listen to THEIR stream, on THEIR dime, in the first place.
It is as fair a system as one can get (with the obvious exception of the aforementioned firewalled users, and an unavoidable portion of people like yourself who, as you say, "sure as hell [are] not going to pay for someone else to get music/videos/pictures/etc at my expense" and will simply refuse to cooperate with the network, both types of whom make things more difficult and expensive for the rest of us with proper connectivity and a fucking spine).
If the cost in your case of providing this service to listeners, and indeed, the originator of the stream, is too great for you to muster, you'll be better off sticking with FM radio and (if applicable) MTV, either of whom who will happily allow you to sit around and consume their material all day, free of charge.
Meanwhile, leave the rest of us alone, you blood-sucking, parasitic tick. You've got the wrong attitude to participate in anything requiring a cooperative effort between peers, where the only cost of admission is that you do what you can to share what you've got.
Re:Who are they trying to recruit? Linux users?
on
Linux Games WIth Guns
·
· Score: 5, Interesting
A few months ago, I went through [most of] basic training at Ft. Benning, in a platoon with a bunch (as in, all but 3) of geeks, though not all of them were Linux geeks. I noticed a strange domination of Cisco yuppies with Novell skillz, harping on about Win2k. Plenty of UNIX programmers, though, which is what matters.
These geeks were in Ft. Benning (which normally only trains infantry) because Ft. Jackson (whom normally trains these sorts) was booked solid.
Thus, I'd like to say that not only is the Army hiring Slashdot-reading linux-using teenagers, but that they're recruiting them in droves.
And, yes, it -did- look like a Linux convention for the first week or so, but they were all looking and -acting- like soldiers within a couple of months. Even the former three-pack-a-day gamers were partaking in nonsensical pushup competitions, and the platoon was top in the company for fitness scores.
It was a strange transition to see, watching the flabby, pale, quiet Linux kids turn into hardcore rope-climbing, gun-toting freaks.
And in a few months, when those same kids are done training at Ft. Gordon and get shipped out to the front line as communications geeks, I'll be very glad to have them there, while I sit on my once-again flabby, civilian ass and read Slashdot.
I'm running XP on a K6-2 350, with 320 megs of RAM, and a bunch of 7200 RPM IBM SCSI disks.
It -feels- faster than Win98SE did, and it's a great deal more stable. Some things are a little strange about it (I have a fourth-party driver for the Voodoo3 3500TV in this machine, for instance, to get all features of it working), but it's been good to me.
Turning off the whiz-bang features helps speed considerably. My start menu doesn't scroll or fade or any of that nonsense, for instance, and it is thus quite responsive. About the only gee-whiz feature I keep around on XP is the funky window dressing, and that doesn't seem to cause any significant performance hit.
Since installing it, I've been using it a good deal more than the Slackware-based Linux install that I've been growing for the past 6 years on the same machine, and find fewer and fewer reasons to boot Linux.
OTOH, I've always got a FreeBSD router/SMB server which gets me my *nix fix, not to mention an old laptop with Slackware 3, and Cygwin installed under XP.
The machine my girlfriend uses for browsing, email, IM and such is a lowly P133, with 96 megs of RAM, running 2k. It also seemed to get a bit zippier in upgrading from 98SE, and I haven't noticed any significant swapfile thrashing those times that I've used it.
That all said, back in the early 90s when I was running Desqview on a 386, Windows 3 was just starting to ship with new machines. I liked the interface well enough, but I hated it because it broke constantly and just plain didn't work most of the time. Awhile later, I was running OS/2, and still thought that Windows' simplicity would be great, if the magic actually worked. Even when I started with Slackware around '96, I kept thinking "Gee, there's all kinds of software for Windows, and this new PCI plug-n-play stuff looks really cool. Too bad the OS is such trash."
It's gotten a lot better. It's good enough for me to use daily -- and it's about fucking time that happened, having been under serious fucking development by a huge fucking corporation for TEN FUCKING YEARS.
Common inkjet printers are very dumb, inexpensive machines. They accept instructions and execute them justabout immediately, with the only logic required on their part being "Gee, do I have enough data in the buffer to print a complete line yet? If so, I might as well print it. If not, I'll wait."
Not much smarts there, except what's required to feed paper correctly, measure out the proper amount of ink, and tend the buffer. They're only a step above the similarly-cheap Star NX-10 9-pin dot matrix I had in 1988 in terms of intelligence.
Contrast this with what's required to implement Postscript interpretation for modern color printing:
A fast CPU to render the print. As resolution increases, this becomes more of a problem - I've got an HP here which prints at something like 2400dpi. Add to this that Postscript is a rather complete programming language, probably capable of, say, printing FFT sonograms of MP3 files being fed to the printer, and the need for a fast CPU grows even more. (Anyone have a Postscript version of seti@home?)
One also needs a huge buffer, preferably one which can hold several pages of rendered data (because the rendering is so -slow-, and you don't want it to render each of 50 copies of a 3-page report individually).
And, the license. I don't suspect this costs more than a dollar or two per device - a price more than made up for by the selling point of supporting Postscript - but it is an additional expense which should be noted.
Fast, dedicated CPUs and capacious memories are simply not within the realm of the $100 inkjet printer. Besides, every modern computer already has a fast CPU and a huge amount of memory - why duplicate that in a printer which is normally only used a few times a week, on average? It's overkill.
HP sells (or at least did a couple of years ago when I was looking) some rather high-end inkjets which support Postscript and plug into the network directly. I never got to try one in real life, because they were deemed too expensive. (A Linux box with Ghostscript currently fills that void.)
Postscript is nice, and very complete and capable and enough of a slow-moving and available target that it's possible to accept it as a standard. If only it weren't so bloody expensive to use...
Back on topic: A standardized way to feed bitmaps to dumb inkjets would be an astoundingly good thing, and free to implement.
Is there some reason why your post reads with the semi-literate, uneven verbosity of the sort not normally found outside of Japanese stereo instruction booklets?
I hope you realize that merely pressing the power button of a modern PC or monitor does not turn it off, but merely put it to sleep. They still draw power, but the fans stop spinning. And a fan that doesn't spin, moves none of the air that the components (which are still quite capable of generating heat) require for cooling.
I was asleep, one night, with my ISP's newly-built DNS server sleeping (ie: in the ATX soft-off state, with the power cord still plugged in) on the floor beside the bed. When I woke up some hours later, it was puking blue smoke from its power supply. Everything survived, except the smouldering PSU, and the box still serves queries (with killer uptimes).
I've never had a computer catch fire while it was operating, however. Thus, my tendancy is to leaving everything running at all times, for fire prevention.;)
On a more serious note, uou should just fucking unplug everything when you're done with it. And don't just plug everything into a power strip and use that as a master switch: I mean, unplug it -all-. Surge protectors are fire hazards, by themselves.
Don't forget to unplug any electronic lighting controls (yes, even the nice Leviton controls in the CEO's office need unwired daily), security lights (can we say superheated, over-charged gel cell batteries?), and pencil sharpeners (the electric motor uses coils, just like a heating element), as well. Oh, and your cell phone charger - do you know that the battery could EXPLODE if the charger were to fail? And none of this is to mention what would happen if, say, a loose connection on a ventillation fan were to arc and start flaming insulation inside a ceiling somewhere, where it'd likely burn for quite awhile before someone noticed.
And be sure to disconnect any telephones, and networking equipmen, and anything else which is connected to something else with copper.
In fact, now that I think about it a bit, it seems that the only way to prevent electrical fire disasters is to unplug the building, and throw the cables across to the other side of the street.
Better do the same with the gas, too. Just to be safe.
And make sure that all company vehicles have their batteries disconnected after use, and are parked off-site. There's always that non-zero chance that the radio will figure out a way to make fireworks from the tiny amount of current used to keep the clock on time.
I pity the first-shift people who have to bundle up extra warm in January while they wait for the building to re-heat after plugging it back in every morning, and especially those who get to put the cars back together every day, but I'm sure that they realize that it's all For The Good of The Company.
What's this talk of minutes and seconds? I thought it was all supposed to be expressed in the form of a simple unit and a multiplier prefix.
A kilogram, for instance, is a measure of 1000 grams. A deciliter is 1/10th of a liter.
It is nonsensical to use conventional naming in a new system such as this, as aside from the obvious ensuing confusion, it fails on the metric system's main claim of fame: easy of conversion between units, and rapid understanding of new/different measurements.
I don't know where "gram" came froom, but I can only assume that it's the name of the person who specified its weight.
That said, I propose a new system of time, based on the Osborne:
There are 10 Osbornes per Earth rotation. Each Osborne is comprised of 10 deciOsbornes. Each deciOsborne is comprised of 10 centiOsborne. Each centiOsborne is comprised of 10 milliOsbornes.
And so on, and so forth. You get the idea.
Which is the whole point, really - it is identifiably similar to the rest of the metric system, and thus easily understood by default. You already know what the prefixes mean, and can apply them instantly.
I'll take such long-hand decimal notation any day over things expressed in minutes and seconds.
...and to think my car tunes itself, as I drive down the highway.
Some stock setups, with catalytic converters, in a state of complete negligence, can still pass Canifornia's smog inspection. Some stock setups can still get 30 miles per gallon of regular unleaded gasoline. Some *modified* setups can crank out upwards of 300 hp from a 2.4 litre four cylinder watercooled engine, drive a car that seats 5 upwards of 150 mph, and 0-60 in under 7 seconds.
My '95 Beretta has a 2.2 liter Chevrolet engine that gets >25 miles per gallon, 0-60 in about 8.5 seconds, and hits its in-built governer hard at 113MPH (I'd take it faster if it'd let me). It doesn't drink coolant, doesn't need chips, I pour oil into it when I feel like it, and it just plain runs. I did replace the (factory) plugs at around 100k miles (the horror!), after it began stumbling on cold, wet mornings. Now it starts within a couple of rotations, even when below 0.
The Stock Chevy engine can be overhauled; rebuilt, for under $500, as with most other single-cam 4-bangers. It does take a team of seven midgets and a step stool to remove it, however.
These cars DO have their limitations, but a lot of us driving modern, bone-stock, base-model economy grocery getters often wonder whether the old codgers who harp on about the beauty of an air-cooled engines left what remained of their brain in 1972 on the head of a pin, stuck to a big hairy resin ball, on the coffee table right next to Bertha (the plumbing store bong), a pile of dusty-looking sandwich bags, and two empty three-peice pull-tab cans of PBR.
There's no problem with reducing or re-arranging the cooling for a CPU. It is not rocket science.
Perhaps it is irrelvent these days, but I have a 133MHz Pentium box that just does web-browsing, e-mail, and seti@home. A couple of years ago, the bearings on its fan got noisy. So I dug around in The Drawer for a couple of minutes looking for socket 7 heatsinks. It's been running with the largest chunk of black aluminum I had handy, without a fan, for years. It is/was also the most stable Win98 machine I've ever seen, going for months at a time between reboots. It's doing just as well now with Win2k.
I'll never have to bother with that CPU cooler again. The only critical moving parts now are the (solitary) PSU fan and hard disk.
Compared to the roar of the ethernet switch, firewall box, and gaming rig across the room it's essentially silent.
That all said, some people here seem to think there's some magic to having a high-speed whiney piss-ant fan directly on the heatsink.
Sure, a modern CPU is likely to be less-than-tolerant of running without active cooling. They make more heat, and have less surface area to spread and conduct that heat away from the core.
But, the heatsink doesn't care. It just wants air passing over/through it, and nearly all of them want that airflow coming from the top.
All fans in this context have very plain, easy-to-understand airflow ratings, expressed in Cubic Feet per Minute. If your current fan moves say, 40CFM, and keeps things sufficiently cool, a larger, quieter replacement and duct which is also capable of moving 40CFM will work just as well.
Yes, the duct impede airflow somewhat. Your CPU may run a degree or two warmer because of it. This, for all purposes, does not matter.
Just make sure it's coupled reasonably well to the heatsink, as it does take a very small amount of pressure to force cool air down between heatsink fins. You'll gain an ounce of efficiency in this way over simply directing air toward the heatsink.
FWIW, Alpha heatsinks are widely available without fans at all. They tend to have high-end qualities and price, but it's probably a drop in the bucket compared to the rest of this project.
Materials for the duct can be almost anything. I'd be inclined, personally, to use 1/16" ABS plastic, but only because I've got a bunch of it sitting around and it's a nice shade of purple.
If you feel like over-engineering things, try to radius (curve) the bends in the duct and promote laminar flow. Or just do a hard 90 degree bend -- at these velocities, it really doesn't much matter.
Other construction materials might include pipe. The plumbing section of a good hardware store will have for sale a plethora of nice white PVC fittings, which have the added benefit of promoting laminar airflow out-of-the-box. It shouldn't be too difficult to come up with some combination which can be bolted to the fan, have a nice, smooth reduction in size, a bend, and then fasten securely to the heatsink.
Or, for a rustic look, one could enlist the services of a someone who makes ducts for a living. A tin smith should be able to bend and cut something like this together in a few moments time, and he'll likely do it cheaply just because it's an interesting project that does not involve a furnace (Athlon XP notwithstanding). Having flanges built-in to drive screws through, into the fan and heatsink, would be secure and trivial to implement. (If I didn't have the ABS handy, I'd investigate this route first.)
Good luck!
Re:currency tracking hardly needs rfids
on
Greenbacks No More
·
· Score: 2
Nice idea, but horribly flawed.
A "unique" number which has not been printed on a legitimate bill (and, yes, the Treasury -already- knows the serial numbers of every bill ever released into circulation) would trivially and automagically be flagged as counterfeit.
Which means, since you can't just make them up and get away with it, that the counterfeiter must borrow serials from legitimate bills.
This leaves open the possibility that two uniquely identical bills (the counterfeit, and its legal clone) are in two places at the same time. Which, with this mythical database, is also trivial to flag.
It's not quite as easy to circumvent as you think.
In a community of 70 homes, the distance capabilities of single-mode fiber as a medium are rather out of context.
Coax can propagate 1GHz signals (read that as "10/100/155(oc3)/1000 and beyond", if you wish) with ease.
Distance, in this application, does not matter. He's not looking to wire the countryside for a ten mile radious, but rather piggyback some bandwidth alongside the new sanitary system in a village of a couple of hundred heads.
Fiber is neat. Per foot, on a reel of a few thousand feet, it's pretty cheap.
Every other aspect of it is painfully expensive. It is not a magic bullet. It does not equate to enormous increases in bandwidth over copper. It does not make sense for any communications need which is not point-to-point and of extremely long distance, except for precious few situations where it is necessary to have no electrical contact between two points, or where radiated or induced noise is intolerable.
I'm a Time Warner customer. We don't, as yet, have any sort of download cap, except for a 2Mbps downstream limit (wah, wah). Their provided news server is similarly free.
Only problem is that I'm having trouble finding decent software to work the magic. There is, of course, Agent and Free Agent, which seem to do an OK job of interactively decoding multipart binaries. But what about the *NIX side of things?
Are there any -good- packages for Linux, or in the FreeBSD ports collection, or available for MacOS X? What's good for pulling down every file in a group? What's good for queueing and performing more selective downloads?
Does software exist with even rudimentary file sorting abilities, so that I could, say, tag a bunch of pictures with random filenames of a girl named Charlotte and move them to their own directory with ease, using the subject line to sort them?
All aspects of fiberoptic networking are expensive. It's expensive to lay, terminate, splice, repair, utilize, and maintain.
Cable companies, these days, have fiber all over their network, but not to people's homes - the "last mile" is still done with coaxial copper wire.
From what I recall of my visit to the new Time Warner building here, my home town of ~30k is split into 6 nodes. Which is to say: Six fiber runs, for thousands of coaxial drops. It works well.
Coax is cheap. It's cheap to lay, cheap to terminate, easy to splice and repair, and networking gear to utilize it is readily available from consumer-oriented companies like Linksys and Toshiba. It can be serviced by mere mortals with tools available from Radio Shack or Wal-Mart. It can carry DC current to power any amplifiers which are needed in-line, while fiber is completely non-conductive. It is more versatile than fiber; I have a number of devices, from my stereo to my cable modem, which can interface directly with the coax network. I have -zero- which have a fiberoptic port. My apartment building is wired for coax, with fiber nowehre to be seen. So on, so forth.
It is also readily able to stretch the relatively small distances encountered in your community with coax. Single-mode fiber is made to go for tens of miles, and for your application, it's just not needed. Add to this the high cost of plugging people in, and you'll soon find that, beautiful though it may be sitting there underground, your fiber network is completely unused.
Just what the world needs. More dark fiber.;)
A coax network will also let you install a community antenna for people to use, in the original spirit of cable TV, in addition to providing the financially-dubious service of internet access to a mere 70 homes. It would be an excellent way to supplement the DirecTV dishes which are undoubtedly common where you live, get rid of ugly roof-mounted antennas, and probably be financially feasible to provide for free (built into property taxes or somesuch).
That all said, if all you want to do is put people online, it'd be so much cheaper and easier and, most likely, better to construct a tower large enough to see the entire community, and spray the area with 802.11b. Wireless ethernet, even properly implemented with cellular-style antenna arrays at the head end, parabolic grid antennas mounted on houses, and high-end Cisco Aironet gear all 'round, is still vastly cheaper than anything involving fiber for the masses.
If, having given that some serious research, you still have your heart set on laying cable, it'll be in the plan's best interest to do something with it which might -actually- get used at some point in time -- which is to say, lay anything but fiber.
Consider burying some 500-pair telephone wire and installing a DSLAM at the fire station or other appropriate public structure. You'll have far better luck offering high-end SDSL service than trying to interest people in spending thousands on fiber gear.
When building a system from scratch I try to buy as much as I can from memman.com. Outstanding customer service, cheap prices, and flat-rate shipping. Lifetime warranty on RAM.
I ordered an Athlon XP and a low-end CPU cooler from them recently. They called me a short while later, and apologized for not having the heatsink I wanted in stock, and then offered to send a rather high-end solid copper cooler instead. No extra cost, of course.
Another time, there was some manner of trouble with the order, and I wasn't aware of it until the next day. It was sorted out quickly by the time I got a chance to call them, but after I hung up I felt a little bummed that the parts would be a day or three later in arriving.
Unbeknownst to me, they upgraded the shipping to 2-day FedEx instead of ground. The stuff showed up on Friday, instead of the next Monday or Tuesday like I expected.
This level of service seems to be par for them, and I've been giving them as much business as possible since I discovered them sometime in the mid-90s, with a cheesy ad in the back of Computer Shopper.
Contrast this with other online vendors, who will sit on your order for several days, silently, and have no idea who you are or what you want when you call to ask where your hardware is -- the people at memman.com seems genuinely interested in keeping customers happy, and are a joy to work with.
I'm beyond pleased with their level of service.
(Nope, I don't work there. I'm not a shareholder. They don't advertise on my porn site. So on, so forth.)
IIRC, Alexa is responsible for the content of Netscape's "What's Related" button, and they've been, appearently, taking snapshots of whatever they could for years. I seem to recall some discussion about this button, and the data-collection policies at Alexa about the same time the button started appearing in Netscape.
Ironically, despite archive.org's extensive cache and slashdot's search feature, I can't find it. Hrmph.
According to whois, archive.org and alexa.com are both registered to companies in San Francisco. Additionally, the 9/11 TV news archive page, as linked from archive.org's main page, credits a number of Alexa employees in the right-hand sidebar.
Slashdot Mirror
You must be looking very hard for the things you seem to want...
RadioShack springs to mind as a place to pick up an optical switch for toslink, fairly cheaply. Whether it uses mirrors, or close-coupled light pipes, or whatever, it probably works justfine.
Midiman sells a converter (called "CO2", IIRC) which converts between optical and coaxial SP/DIF. They pop up on Ebay from time to time, and work well for a variety of tasks (SCMS stripping, anyone?).
The inexpensive Audio Alchemy DDE v1.1 that I use as a DAC for my computer has coaxial and optical inputs, as well as an isolated, re-clocked coax output which just mirrors whatever input is selected.
Edirol has a product here which has five connections: USB, toslink in/out, and coaxial in/out.
Finally, as configurable as the sound support is on the Shuttle board, I'd be very surprised if one couldn't turn one of its three multi-purpose jacks into a coax output using nothing but a software switch.
Back In The Day, I made an informal study of this topic, on IRC (of course).
I chatted with prospective participants, gathered as much information from them about their modem as possible, asked if it were OK if they might drop carrier, and sent three magical ICMPs their way, one second apart. I logged the data in a text file, which I don't seem to be able to find just now.
Something like 80% were vulnerable. I was amazed. I gave my SupraFAXModem 288i a big fat kiss for being such a good sport about it all.
Hacks like this have existed for a -long- time. I remember helping sysops in 1991 to defend against kids who would dial up BBSs, and use the login prompt's echo to drop the board's modem to command mode, whereupon they'd leave it in a strange state and things wouldn't work until the sysop reset the modem.
Fun stuff. It wasn't until some years later that I learned about the Hayes patent on having a silence period between +++ and command mode.
As some other posters have pointed out, you don't need to have BIOS-level access to the entire hard drive to get the machine to boot properly and run Linux. Just keep the bootable partition within the first 500 megs, and all is dandy.
However, you might want to use a BIOS overlay, anyway. Makes things easier, and lets you do stuff like run ancient DOS games on the hardware they were meant for, or fire up OS/2 Warp for an old-school look at the future.
On my 386SL-25 laptop, I'm using IBM's overlay software, which is freely downloadable as a bootable floppy disk image. Most other manufacturers also supply overlay software, free, but it generally requires you to install it on a machine with a drive of the same make; Maxtor's software needs to see a Maxtor drive somewhere in the system, or it will simply refuse to cooperate.
This works fine with Linux, and has for a very long time. It just recognizes that the drive was partitioned with overlay software, and does the same sector translation on its own.
I doubt there's even a speed hit.
I used IBM's software because I have a bunch of their SCSI disks in the machine I was borrowing to do the Linux install with, and Hitachi stands almost alone as a vendor who doesn't supply overlay software of their own. (the laptop, sadly, has no floppy drive, CD-ROM, or other external storage, so I spent most of an evening swapping IDE cables trying to get the thing to boot.)
The component most likely to complain about temperature and vibration is the hard drive.
:-/
/dev/zero > /dev/null, or whatever) to keep the motherboard components warm and well within design spec, which will help reduce condensation. Do the opposite on hot days - HLT the CPU when idle, so as not to make things any warmer than needed to get the job done, and take whatever other power-saving measures you can.
While your car's suspension will work rather well at isolating things from bumps in the road, it wouldn't hurt to get a little creative with mounting the drive. Suspending it between rubber straps might be a good idea, but watch out for things which are bouncy (resonant) -- it almost doesn't matter how much of the initial impact that your suspension absorbs, if it continues to shake the hell out of the thing several seconds after stuff should've come to a rest.
Hard drives come packed in open-cell foam, like a mattress pad, and seem to survive UPS Ground pretty well in such an arrangement. It wouldn't take too much creative engineering to fabricate an enclosure made predominately out of foam, but with a fan and enough room for air to circulate.
Temperature doesn't look like it's too much of a hassle, these days. I'm looking at the specs on an IBM 120GXP, which show it to be happy from -40 to 55 degrees celsius (-40 to 131 F). Since these are ambient temperatures, the drive is thus designed to withstand a trunk at 131 degrees with moderate airflow.
Since you'll be opening your windows and/or turning on the air conditioning Right Away on such blazing hot days, the hard drive should start recieving cooler ambient air at about the same time it starts generating appreciable heat of its own.
And in the winter, at -40, the last thing you're worried about is whether your hard drive will spin up. More important is whether or not the engine will, and if you'll be able to get the tires un-frozen from the ground. If you've got half a brain about you, you'll have the computer and stereo off, anyway, until the car is well under its own power...and, by that point, generating heat to warm up the electronics.
I'd avoid connecting the enclosure directly to the car's ductwork. Ever see a windshield fog up on the inside on a cold morning, just after you turn on the defroster? Imagine that happening to your in-car PC. Bad news.
Since the motherboard you've chosen is so bloody small, have you considered putting it under the front seat? Things would get impossibly tight in a sports car with power seats, but should be do-able in almost anything else. There's no reason for the box to be more than an inch or three high. Sescom is a company who makes a large variety of metal boxes for do-it-yourself projects - chances are, one of them would fit your motherboard, hard drive, and some manner of DC power supply justabout perfectly, while remaining small enough to slide under the seat.
Drill or punch holes in the sides of the box, and mount the biggest, lowest-RPM fan you can find on the top of the box, blowing down. You might even be able to do away with having a dedicated CPU fan, and get by with a just large heatsink.
There's a few other things you might want to look into, with software. There's a way to poll hard drive temperature using a protocol called SMART - if the drive is cold, keep it spun up to reduce the viscosity of the grease in its bearings. Likewise, if the drive is fairly warm, spin it down when not needed to reduce wear and help ward off data loss (they are, of course, -much- more durable with not spinning).
Do the same with the CPU, if it suits you. If it's cold out, run the CPU full-tilt (seti@home, some random busyloop, cat
The goal here is to bring the board to some desired operating temperature, and keep it there until the car turns off.
All that said, you'll probably find that the most sensitive component of your system is the LCD display, which will be painfully slow/frozen on cold mornings, and either solid white or black after a hot day of sitting in the sun...and there's really not much to be done about it, unfortunately.
good luck!
A 64x CAV CD-ROM drive will spin at a maximum of 12,800 RPM, according to my math, but drives don't seem to exist at that speed just yet. More common 56x drives spin at 11,200 RPM.
If things are consistantly and violently exploding at a little over twice that speed, would it not make sense that there is a very real safety issue in making things faster than this?
One might theorize that off-balance discs, cracked/scored/otherwise-damaged media, and just plain bad luck might cause things to go dangerously amiss even at current speeds.
Current high-end SCSI hard drives spin at 15,000RPM, but do so using extremely well-balanced, carefully-produced, expensive solid aluminum platters and motors. And, besides, they're also encased in heavy metal boxes, and don't have a soft plastic face through which to fire shrapnel into the chest of the user.
Consider that a CD-ROM has a much larger diameter than a typical hard disk platter, and is thus exposed to far greater centrifugal force and linear velocity. Consider also that a CD-ROM drive only costs a few dollars to make, and that CDs are down to a couple of cents each in large volume.
Given this information and that contained in the article, I doubt it would take much effort to make a CD explode in a current 56x drive, thus presenting a very real bottleneck, indeed.
Plextor drives can be 'underclocked' with software switches fairly trivially, using free tools. They also let you do some fanciness like control spin-down times, and how the drive responds to errors.
OTOH, the fastest drive Plextor makes operates at 40x CAV. And I've fed many, many messed up (cracked, deeply scratched, off-balance) CDs into my 32x Plextor, without ever having one disintegrate, even when they're spinning at high speed for 12 hours or more.
So, personally, I'm not too worried about the safety aspects. But if you want to slow down a CD-ROM for whatever reason, Plextor is a very sure route to follow.
Scary prospects, indeed, but not much (if any) worse than the current situation. There's a number of mechanical components in conventional vehicles with no redundancy and a tendancy to fail now-and-then.
If a tie rod snaps on a 1985 Buick LeSabre, you're quite simply fucked, unless you happen to be driving slow, in light traffic, or sitting at a stop sign or something.
If the rear hub assembly (think "spindle" or "wheel bearing") snaps in two on a 1995 Chevy Beretta, you're also probably fucked. The body bottoms out on pavement, the car turns around in a quick 270, and tends to disarm any attempt at serious braking. Half of one side of the rear brakes end up scattered along the roadway, and the other half stay bolted to the wheel as it rolls away and bounces off of a guard rail. This completely disables to the mechanical brake, and makes the hydraulic brake largely ineffective as fluid pours out of the smashed brake piston -- the pedal goes to the floor with remarkable ease.
Or, take ball joints as another example. Wheel ends up flopping freely around, and the strut now entrusted with keeping it attached gets ripped apart. Steering is iffy because one of the front wheels is locked hard right and smoking, power is gone (the half shaft gets ripped out of the transmission), and there's a good chance that a brake line will be snapped, making stopping rather interesting. (1993 Mercury Villager.)
And, of course, tires go flat. They explode, they remove themselves from the rim, they delaminate - and those not mounted on Ford Explorers are not exempt from this. I tore one off of the rim on the aforementioned Beretta while going sideways through a ditch, into a field of corn stubble at 80MPH (yep, that one was my fault).
This stuff happens, it's usually fairly spectacular, very fast, and there's no redundant parts in an automotive suspension or steering system to ensure continued control of the forward progression of the vehicle when these components fail.
Interestingly, the one fly-by-wire mechanism in that Chevy (the fuel injection system) has never shown any signs of ill intent.
Given that last point, and the failure rate of mechanical parts, I have no problem entrusting my life to a simpler system of electronic devices. That car has been through numerous water pumps, a couple of alternators, and various suspension parts (some more than once), but I've never had issue with any electrical system it has.
It is also worth noting that the power brakes and steering you blame for your inherent mistrust are also strictly mechanical systems...
HTH. HAND.
I have no numbers to back this up, but in my experience, 2.4 -feels- slower on older hardware than 2.0.
;)
'Sides, unless it's a router box and you need the latest, greatest QoS tools and security fixes on your 486, there's very little reason to upgrade.
Most people seem to get new kernels when they need support for new hardware, and that's just not much of an issue for a 486.
I did upgrade the kernel on my 386SL/25 laptop recently, from 2.0.37 to 2.2.18, but only because I wanted to play with the swsusp patches and didn't feel like learning how to backport them to 2.0.
The temptation to move to a 2.4-AC kernel with swsusp built-in was very easy to resist.
Ah, but you forget that SCSI is backward-compatible. Which is to say that every SCSI card is a SCSI-1 card. Even cards supporting Ultra320, capable of transferring 320 megabytes per second, using balanced signalling at 160MHz on a bus 16 bits wide will happily and efficiently step down to the 5MHz, 8-bit single-ended bus most people call SCSI-1, as long as things stay terminated properly.
This is also to say that a 15KRPM Ultra320 SCSI disk will happily work on a 15-year-old Mac, or an Amiga or Atari or IBM PC(tm) with some manner of SCSI interface, given only the correct adapters and termination.
It makes things fun.
Bah. Real men don't use backups.
In instances where fsck begins whining about things being seriously wrong, I politely tell it to fuck off. The machine then boots up with the partition in question mounted read-only.
This gives me a shot at doing a quick backup of any Really Important Stuff to tape, CD, or just another drive.
This hurts nothing in the process, but does cost a bit of time.
After that's done, I'll tell fsck to go ahead and smash the drive. It usually works great, at least on Linux, but now and then has issues which would've trashed my Really Important Stuff.
No big thing, though. Restore the Really Important Stuff from the freshly-made backup, and rebuild the rest of the filesystem by hand.
In a properly-implemented system of this type, each listener would provide a stream for n additional listeners, only when actually listening (where n is a user selected integer of fairly small value). If n=>~3, one would be able to help out the fraction of listeners who are hopelessly and forever firewalled and thus unable to mirror the stream themselves.
At any rate, you turn off the "radio" and go to bed - and all bandwidth consumption halts until such a time as you feel like turning it on again.
You shouldn't look at it as paying for someone else's music - that someone may very well be in the same boat as you, and must pay seperately to -recieve- the bandwidth you give them. Not to mention the poor soul who was kind enough to let you listen to THEIR stream, on THEIR dime, in the first place.
It is as fair a system as one can get (with the obvious exception of the aforementioned firewalled users, and an unavoidable portion of people like yourself who, as you say, "sure as hell [are] not going to pay for someone else to get music/videos/pictures/etc at my expense" and will simply refuse to cooperate with the network, both types of whom make things more difficult and expensive for the rest of us with proper connectivity and a fucking spine).
If the cost in your case of providing this service to listeners, and indeed, the originator of the stream, is too great for you to muster, you'll be better off sticking with FM radio and (if applicable) MTV, either of whom who will happily allow you to sit around and consume their material all day, free of charge.
Meanwhile, leave the rest of us alone, you blood-sucking, parasitic tick. You've got the wrong attitude to participate in anything requiring a cooperative effort between peers, where the only cost of admission is that you do what you can to share what you've got.
A few months ago, I went through [most of] basic training at Ft. Benning, in a platoon with a bunch (as in, all but 3) of geeks, though not all of them were Linux geeks. I noticed a strange domination of Cisco yuppies with Novell skillz, harping on about Win2k. Plenty of UNIX programmers, though, which is what matters.
These geeks were in Ft. Benning (which normally only trains infantry) because Ft. Jackson (whom normally trains these sorts) was booked solid.
Thus, I'd like to say that not only is the Army hiring Slashdot-reading linux-using teenagers, but that they're recruiting them in droves.
And, yes, it -did- look like a Linux convention for the first week or so, but they were all looking and -acting- like soldiers within a couple of months. Even the former three-pack-a-day gamers were partaking in nonsensical pushup competitions, and the platoon was top in the company for fitness scores.
It was a strange transition to see, watching the flabby, pale, quiet Linux kids turn into hardcore rope-climbing, gun-toting freaks.
And in a few months, when those same kids are done training at Ft. Gordon and get shipped out to the front line as communications geeks, I'll be very glad to have them there, while I sit on my once-again flabby, civilian ass and read Slashdot.
I'm running XP on a K6-2 350, with 320 megs of RAM, and a bunch of 7200 RPM IBM SCSI disks.
It -feels- faster than Win98SE did, and it's a great deal more stable. Some things are a little strange about it (I have a fourth-party driver for the Voodoo3 3500TV in this machine, for instance, to get all features of it working), but it's been good to me.
Turning off the whiz-bang features helps speed considerably. My start menu doesn't scroll or fade or any of that nonsense, for instance, and it is thus quite responsive. About the only gee-whiz feature I keep around on XP is the funky window dressing, and that doesn't seem to cause any significant performance hit.
Since installing it, I've been using it a good deal more than the Slackware-based Linux install that I've been growing for the past 6 years on the same machine, and find fewer and fewer reasons to boot Linux.
OTOH, I've always got a FreeBSD router/SMB server which gets me my *nix fix, not to mention an old laptop with Slackware 3, and Cygwin installed under XP.
The machine my girlfriend uses for browsing, email, IM and such is a lowly P133, with 96 megs of RAM, running 2k. It also seemed to get a bit zippier in upgrading from 98SE, and I haven't noticed any significant swapfile thrashing those times that I've used it.
That all said, back in the early 90s when I was running Desqview on a 386, Windows 3 was just starting to ship with new machines. I liked the interface well enough, but I hated it because it broke constantly and just plain didn't work most of the time. Awhile later, I was running OS/2, and still thought that Windows' simplicity would be great, if the magic actually worked. Even when I started with Slackware around '96, I kept thinking "Gee, there's all kinds of software for Windows, and this new PCI plug-n-play stuff looks really cool. Too bad the OS is such trash."
It's gotten a lot better. It's good enough for me to use daily -- and it's about fucking time that happened, having been under serious fucking development by a huge fucking corporation for TEN FUCKING YEARS.
Postscript is expensive.
Not to license, but to implement.
Common inkjet printers are very dumb, inexpensive machines. They accept instructions and execute them justabout immediately, with the only logic required on their part being "Gee, do I have enough data in the buffer to print a complete line yet? If so, I might as well print it. If not, I'll wait."
Not much smarts there, except what's required to feed paper correctly, measure out the proper amount of ink, and tend the buffer. They're only a step above the similarly-cheap Star NX-10 9-pin dot matrix I had in 1988 in terms of intelligence.
Contrast this with what's required to implement Postscript interpretation for modern color printing:
A fast CPU to render the print. As resolution increases, this becomes more of a problem - I've got an HP here which prints at something like 2400dpi. Add to this that Postscript is a rather complete programming language, probably capable of, say, printing FFT sonograms of MP3 files being fed to the printer, and the need for a fast CPU grows even more. (Anyone have a Postscript version of seti@home?)
One also needs a huge buffer, preferably one which can hold several pages of rendered data (because the rendering is so -slow-, and you don't want it to render each of 50 copies of a 3-page report individually).
And, the license. I don't suspect this costs more than a dollar or two per device - a price more than made up for by the selling point of supporting Postscript - but it is an additional expense which should be noted.
Fast, dedicated CPUs and capacious memories are simply not within the realm of the $100 inkjet printer. Besides, every modern computer already has a fast CPU and a huge amount of memory - why duplicate that in a printer which is normally only used a few times a week, on average? It's overkill.
HP sells (or at least did a couple of years ago when I was looking) some rather high-end inkjets which support Postscript and plug into the network directly. I never got to try one in real life, because they were deemed too expensive. (A Linux box with Ghostscript currently fills that void.)
Postscript is nice, and very complete and capable and enough of a slow-moving and available target that it's possible to accept it as a standard. If only it weren't so bloody expensive to use...
Back on topic: A standardized way to feed bitmaps to dumb inkjets would be an astoundingly good thing, and free to implement.
Is there some reason why your post reads with the semi-literate, uneven verbosity of the sort not normally found outside of Japanese stereo instruction booklets?
Just curious.
I hope you realize that merely pressing the power button of a modern PC or monitor does not turn it off, but merely put it to sleep. They still draw power, but the fans stop spinning. And a fan that doesn't spin, moves none of the air that the components (which are still quite capable of generating heat) require for cooling.
;)
I was asleep, one night, with my ISP's newly-built DNS server sleeping (ie: in the ATX soft-off state, with the power cord still plugged in) on the floor beside the bed. When I woke up some hours later, it was puking blue smoke from its power supply. Everything survived, except the smouldering PSU, and the box still serves queries (with killer uptimes).
I've never had a computer catch fire while it was operating, however. Thus, my tendancy is to leaving everything running at all times, for fire prevention.
On a more serious note, uou should just fucking unplug everything when you're done with it. And don't just plug everything into a power strip and use that as a master switch: I mean, unplug it -all-. Surge protectors are fire hazards, by themselves.
Don't forget to unplug any electronic lighting controls (yes, even the nice Leviton controls in the CEO's office need unwired daily), security lights (can we say superheated, over-charged gel cell batteries?), and pencil sharpeners (the electric motor uses coils, just like a heating element), as well. Oh, and your cell phone charger - do you know that the battery could EXPLODE if the charger were to fail? And none of this is to mention what would happen if, say, a loose connection on a ventillation fan were to arc and start flaming insulation inside a ceiling somewhere, where it'd likely burn for quite awhile before someone noticed.
And be sure to disconnect any telephones, and networking equipmen, and anything else which is connected to something else with copper.
In fact, now that I think about it a bit, it seems that the only way to prevent electrical fire disasters is to unplug the building, and throw the cables across to the other side of the street.
Better do the same with the gas, too. Just to be safe.
And make sure that all company vehicles have their batteries disconnected after use, and are parked off-site. There's always that non-zero chance that the radio will figure out a way to make fireworks from the tiny amount of current used to keep the clock on time.
I pity the first-shift people who have to bundle up extra warm in January while they wait for the building to re-heat after plugging it back in every morning, and especially those who get to put the cars back together every day, but I'm sure that they realize that it's all For The Good of The Company.
That doesn't sound like the metric I know.
What's this talk of minutes and seconds? I thought it was all supposed to be expressed in the form of a simple unit and a multiplier prefix.
A kilogram, for instance, is a measure of 1000 grams. A deciliter is 1/10th of a liter.
It is nonsensical to use conventional naming in a new system such as this, as aside from the obvious ensuing confusion, it fails on the metric system's main claim of fame: easy of conversion between units, and rapid understanding of new/different measurements.
I don't know where "gram" came froom, but I can only assume that it's the name of the person who specified its weight.
That said, I propose a new system of time, based on the Osborne:
There are 10 Osbornes per Earth rotation.
Each Osborne is comprised of 10 deciOsbornes.
Each deciOsborne is comprised of 10 centiOsborne.
Each centiOsborne is comprised of 10 milliOsbornes.
And so on, and so forth. You get the idea.
Which is the whole point, really - it is identifiably similar to the rest of the metric system, and thus easily understood by default.
You already know what the prefixes mean, and can apply them instantly.
I'll take such long-hand decimal notation any day over things expressed in minutes and seconds.
...and to think my car tunes itself, as I drive down the highway.
Some stock setups, with catalytic converters, in a state of complete negligence, can still pass Canifornia's smog inspection.
Some stock setups can still get 30 miles per gallon of regular unleaded gasoline.
Some *modified* setups can crank out upwards of 300 hp from a 2.4 litre four cylinder watercooled engine, drive a car that seats 5 upwards of 150 mph, and 0-60 in under 7 seconds.
My '95 Beretta has a 2.2 liter Chevrolet engine that gets >25 miles per gallon, 0-60 in about 8.5 seconds, and hits its in-built governer hard at 113MPH (I'd take it faster if it'd let me). It doesn't drink coolant, doesn't need chips, I pour oil into it when I feel like it, and it just plain runs. I did replace the (factory) plugs at around 100k miles (the horror!), after it began stumbling on cold, wet mornings. Now it starts within a couple of rotations, even when below 0.
The Stock Chevy engine can be overhauled; rebuilt, for under $500, as with most other single-cam 4-bangers. It does take a team of seven midgets and a step stool to remove it, however.
These cars DO have their limitations, but a lot of us driving modern, bone-stock, base-model economy grocery getters often wonder whether the old codgers who harp on about the beauty of an air-cooled engines left what remained of their brain in 1972 on the head of a pin, stuck to a big hairy resin ball, on the coffee table right next to Bertha (the plumbing store bong), a pile of dusty-looking sandwich bags, and two empty three-peice pull-tab cans of PBR.
There's no problem with reducing or re-arranging the cooling for a CPU. It is not rocket science.
Perhaps it is irrelvent these days, but I have a 133MHz Pentium box that just does web-browsing, e-mail, and seti@home. A couple of years ago, the bearings on its fan got noisy. So I dug around in The Drawer for a couple of minutes looking for socket 7 heatsinks. It's been running with the largest chunk of black aluminum I had handy, without a fan, for years. It is/was also the most stable Win98 machine I've ever seen, going for months at a time between reboots. It's doing just as well now with Win2k.
I'll never have to bother with that CPU cooler again. The only critical moving parts now are the (solitary) PSU fan and hard disk.
Compared to the roar of the ethernet switch, firewall box, and gaming rig across the room it's essentially silent.
That all said, some people here seem to think there's some magic to having a high-speed whiney piss-ant fan directly on the heatsink.
Sure, a modern CPU is likely to be less-than-tolerant of running without active cooling. They make more heat, and have less surface area to spread and conduct that heat away from the core.
But, the heatsink doesn't care. It just wants air passing over/through it, and nearly all of them want that airflow coming from the top.
All fans in this context have very plain, easy-to-understand airflow ratings, expressed in Cubic Feet per Minute. If your current fan moves say, 40CFM, and keeps things sufficiently cool, a larger, quieter replacement and duct which is also capable of moving 40CFM will work just as well.
Yes, the duct impede airflow somewhat. Your CPU may run a degree or two warmer because of it. This, for all purposes, does not matter.
Just make sure it's coupled reasonably well to the heatsink, as it does take a very small amount of pressure to force cool air down between heatsink fins. You'll gain an ounce of efficiency in this way over simply directing air toward the heatsink.
FWIW, Alpha heatsinks are widely available without fans at all. They tend to have high-end qualities and price, but it's probably a drop in the bucket compared to the rest of this project.
Materials for the duct can be almost anything. I'd be inclined, personally, to use 1/16" ABS plastic, but only because I've got a bunch of it sitting around and it's a nice shade of purple.
If you feel like over-engineering things, try to radius (curve) the bends in the duct and promote laminar flow. Or just do a hard 90 degree bend -- at these velocities, it really doesn't much matter.
Other construction materials might include pipe. The plumbing section of a good hardware store will have for sale a plethora of nice white PVC fittings, which have the added benefit of promoting laminar airflow out-of-the-box. It shouldn't be too difficult to come up with some combination which can be bolted to the fan, have a nice, smooth reduction in size, a bend, and then fasten securely to the heatsink.
Or, for a rustic look, one could enlist the services of a someone who makes ducts for a living. A tin smith should be able to bend and cut something like this together in a few moments time, and he'll likely do it cheaply just because it's an interesting project that does not involve a furnace (Athlon XP notwithstanding). Having flanges built-in to drive screws through, into the fan and heatsink, would be secure and trivial to implement. (If I didn't have the ABS handy, I'd investigate this route first.)
Good luck!
Nice idea, but horribly flawed.
A "unique" number which has not been printed on a legitimate bill (and, yes, the Treasury -already- knows the serial numbers of every bill ever released into circulation) would trivially and automagically be flagged as counterfeit.
Which means, since you can't just make them up and get away with it, that the counterfeiter must borrow serials from legitimate bills.
This leaves open the possibility that two uniquely identical bills (the counterfeit, and its legal clone) are in two places at the same time. Which, with this mythical database, is also trivial to flag.
It's not quite as easy to circumvent as you think.
In a community of 70 homes, the distance capabilities of single-mode fiber as a medium are rather out of context.
Coax can propagate 1GHz signals (read that as "10/100/155(oc3)/1000 and beyond", if you wish) with ease.
Distance, in this application, does not matter. He's not looking to wire the countryside for a ten mile radious, but rather piggyback some bandwidth alongside the new sanitary system in a village of a couple of hundred heads.
Fiber is neat. Per foot, on a reel of a few thousand feet, it's pretty cheap.
Every other aspect of it is painfully expensive. It is not a magic bullet. It does not equate to enormous increases in bandwidth over copper. It does not make sense for any communications need which is not point-to-point and of extremely long distance, except for precious few situations where it is necessary to have no electrical contact between two points, or where radiated or induced noise is intolerable.
I'm a Time Warner customer. We don't, as yet, have any sort of download cap, except for a 2Mbps downstream limit (wah, wah). Their provided news server is similarly free.
Only problem is that I'm having trouble finding decent software to work the magic. There is, of course, Agent and Free Agent, which seem to do an OK job of interactively decoding multipart binaries. But what about the *NIX side of things?
Are there any -good- packages for Linux, or in the FreeBSD ports collection, or available for MacOS X? What's good for pulling down every file in a group? What's good for queueing and performing more selective downloads?
Does software exist with even rudimentary file sorting abilities, so that I could, say, tag a bunch of pictures with random filenames of a girl named Charlotte and move them to their own directory with ease, using the subject line to sort them?
My burner just isn't busy enough, these days...
All aspects of fiberoptic networking are expensive. It's expensive to lay, terminate, splice, repair, utilize, and maintain.
;)
Cable companies, these days, have fiber all over their network, but not to people's homes - the "last mile" is still done with coaxial copper wire.
From what I recall of my visit to the new Time Warner building here, my home town of ~30k is split into 6 nodes. Which is to say: Six fiber runs, for thousands of coaxial drops. It works well.
Coax is cheap. It's cheap to lay, cheap to terminate, easy to splice and repair, and networking gear to utilize it is readily available from consumer-oriented companies like Linksys and Toshiba. It can be serviced by mere mortals with tools available from Radio Shack or Wal-Mart. It can carry DC current to power any amplifiers which are needed in-line, while fiber is completely non-conductive. It is more versatile than fiber; I have a number of devices, from my stereo to my cable modem, which can interface directly with the coax network. I have -zero- which have a fiberoptic port. My apartment building is wired for coax, with fiber nowehre to be seen. So on, so forth.
It is also readily able to stretch the relatively small distances encountered in your community with coax. Single-mode fiber is made to go for tens of miles, and for your application, it's just not needed. Add to this the high cost of plugging people in, and you'll soon find that, beautiful though it may be sitting there underground, your fiber network is completely unused.
Just what the world needs. More dark fiber.
A coax network will also let you install a community antenna for people to use, in the original spirit of cable TV, in addition to providing the financially-dubious service of internet access to a mere 70 homes. It would be an excellent way to supplement the DirecTV dishes which are undoubtedly common where you live, get rid of ugly roof-mounted antennas, and probably be financially feasible to provide for free (built into property taxes or somesuch).
That all said, if all you want to do is put people online, it'd be so much cheaper and easier and, most likely, better to construct a tower large enough to see the entire community, and spray the area with 802.11b. Wireless ethernet, even properly implemented with cellular-style antenna arrays at the head end, parabolic grid antennas mounted on houses, and high-end Cisco Aironet gear all 'round, is still vastly cheaper than anything involving fiber for the masses.
If, having given that some serious research, you still have your heart set on laying cable, it'll be in the plan's best interest to do something with it which might -actually- get used at some point in time -- which is to say, lay anything but fiber.
Consider burying some 500-pair telephone wire and installing a DSLAM at the fire station or other appropriate public structure. You'll have far better luck offering high-end SDSL service than trying to interest people in spending thousands on fiber gear.
When building a system from scratch I try to buy as much as I can from memman.com. Outstanding customer service, cheap prices, and flat-rate shipping. Lifetime warranty on RAM.
I ordered an Athlon XP and a low-end CPU cooler from them recently. They called me a short while later, and apologized for not having the heatsink I wanted in stock, and then offered to send a rather high-end solid copper cooler instead. No extra cost, of course.
Another time, there was some manner of trouble with the order, and I wasn't aware of it until the next day. It was sorted out quickly by the time I got a chance to call them, but after I hung up I felt a little bummed that the parts would be a day or three later in arriving.
Unbeknownst to me, they upgraded the shipping to 2-day FedEx instead of ground. The stuff showed up on Friday, instead of the next Monday or Tuesday like I expected.
This level of service seems to be par for them, and I've been giving them as much business as possible since I discovered them sometime in the mid-90s, with a cheesy ad in the back of Computer Shopper.
Contrast this with other online vendors, who will sit on your order for several days, silently, and have no idea who you are or what you want when you call to ask where your hardware is -- the people at memman.com seems genuinely interested in keeping customers happy, and are a joy to work with.
I'm beyond pleased with their level of service.
(Nope, I don't work there. I'm not a shareholder. They don't advertise on my porn site. So on, so forth.)
IIRC, Alexa is responsible for the content of Netscape's "What's Related" button, and they've been, appearently, taking snapshots of whatever they could for years. I seem to recall some discussion about this button, and the data-collection policies at Alexa about the same time the button started appearing in Netscape.
h tm l/
Ironically, despite archive.org's extensive cache and slashdot's search feature, I can't find it. Hrmph.
According to whois, archive.org and alexa.com are both registered to companies in San Francisco. Additionally, the 9/11 TV news archive page, as linked from archive.org's main page, credits a number of Alexa employees in the right-hand sidebar.
http://tvnews3.televisionarchive.org/tvarchive/
I'd say they're all the same people, more or less. Different corporations, perhaps, but at least the same faces.