How does that stop someone reading the private key from the RFID device using non-RF methods?
You mean like Ross Anderson's many successful attacks on smart cards, using microscopes, logic probes, power-fault attacks, and other really clever methods?
I'm guessing that'll be prevented mostly by 'lawyers'. "Clone this chip and we'll sue your asses, prosecute you under the DMCA, tell the RIAA you've been downloading Metallica, and unplug your freezer until your steaks thaw and make a mess all over your frozen waffles."
Now that you mention it, I've actually done that. I discovered (the hard way) that mineral deposits built up inside the valve get dislodged when it's triggered, and sometimes those chunks then get in the way of sealing the valve again. Which meant a stream of hot water that couldn't be shut off until I shut off the house valve and couldn't be turned on again until I bought a new pressure relief valve.
...will still pale in comparison to ILM. LOTR had some of the worst compositing I had ever seen. The fact that ROTK won an Oscar for Visual Effects only further illustrates the irrelevance of the politics behind the Oscars.
I don't know, I don't think ILM did a great job with Naboo or Tattooine in episodes 1-3, either. The battle of the Gungans vs the robots was a giant load of CGI all around. And that was released in 2005, while ROTK was released in 2003.
But yeah, Gollum was too aggressive for Weta, at least when they filmed it. And although the character was certainly more repulsive, Jar-jar was better done than Gollum (slightly.)
Maybe the new render farm will let them try some better algorithms or more complex models, rather than just speed up the process. We can hope.
I'm of a mixed mindset when it comes to water-cooled datacenters.
On one hand, you've got the makings of a biblical scale disaster with all that water and electricity mixing.
We had such a disaster many years ago.
The coolant lines themselves circled the walls near the ceiling of the dinosaur pen. Beneath each line was a drip tray that was alarmed to sound in case moisture was detected. These drip trays ran the entire length of the coolant lines. That is, the entire length except for about six inches in the very corner, where it was too hard for the tray installers to get to because they were behind the conduit leading down to the main power transformer. I can only assume that all that conduit also made it difficult for the plumbers to properly solder the elbow in the cooling line as well.
Of course there was only one place for the pipe to leak. The version of the tale I heard implies an impressive display of fireworks was seen by all present as the coolant entered the transformer.
Fortunately, the loss of power did not damage the mainframes (except for their ability to run.) The rooftop generator was fired up, and in short order the mainframes were back on line.
And in short order the generator engines stopped because the fuel tanks were kept almost empty, the plan being to fully fuel them only in case of need.
Once the replacement fuel was delivered, the generators ran for only a few minutes before dying again, this time for good. The ancient fuel had congealed in the injectors once the engines had stopped running.
We ran our data center for two weeks powered from a truck-based generator parked in the street while the electricians replaced the transformers and repaired the generators.
I now hear the diesel generator being started every month or two, and run for a few hours.
Rather than a biblical disaster, I'd say it was more like a Marx Brothers' movie.
Ross Anderson at Cambridge has done some great work on the topic. There are many attacks that are documented, and I'm sure there are others. He's used a microprobing workstation to directly access chip features. They've used optical microscopes to view the state of fusible memory cells (such as you might find in a PROM.) This lets them directly read the contents of a chip. To expose the chip, they typically grind through the outer package, or use fuming nitric acid to dissolve the case.
Other, more clever attacks include active testing, where they attempt to get a functioning chip to reveal secrets while it's running. The general idea with this is that you set up a test bench that repeatedly uses the chip to encrypt a known input, change something on the chip (called "fuzzing"), and then check the output is always the same. They've used X-rays to fuzz memory locations. The X-ray beam is used to alter a memory location on the chip -- it almost doesn't matter what spot, really, as long as it causes an error in the output. You then move the X-ray beam to another spot on the chip, and repeat. The idea is that corruptions in a block cypher based on the principles of key rounds (such as DES, AES, Blowfish, Twofish, etc.) will manifest themselves as bits of key material if they corrupt the right round at the right time.
This same sort of principle was used by dropping the chip power at the right time in the encryption algorithm. By causing an error during the encryption process, that error can identify bits of the key.
It's up to you to decide if operating a micro probing workstation, building a precision guided x-ray machine, or bathing chips in fuming nitric acid is "difficult". As far as I'm concerned, yes, I think it's difficult, but I don't have access to that kind of equipment. You might have all of that down the hallway, for all I know.
Hardware advances mostly for games, media and business needs.
Actually, AC, while I don't think you are absolutely correct, you may be on to something there. It's widely believed that many of the advances in home electronics, home theater, computing and networking were due to porn more than any other factor. So if we use that as a starting base, perhaps Chrome was created for a different reason. Maybe it is really destined to be the Ultimate Porn Surfing Engine.
Just think: It'll start small. Google will use Chrome compatibility to partner with porn web operators to offer to protect their site content, and securely ensure payments. It'll work great, and soon all porn sites jump ship and start relying on the Chrome browser. Porn will no longer be viewable on IE or FF, so the world switches completely to Chrome. At that point, Google knows they have the entire internet by its collective short and curlies (almost literally), and that's when they SQUEEZE.
God help us all, we've uncovered it: Chrome's really a plot to take over the world!
It's been real for twenty years or more. IBM has long made a special cash register printer called the Fiscal Printer to meet with Italian sales tax regulations. (Other countries have since adopted the Fiscal Printer standards, but I think Italy was the first.) You can read the programming guide here: ftp://ftp.software.ibm.com/software/retail/pubs/hw/4610/3station/fiscal/italy/fit90n16.pdf
At its heart the Fiscal Printer has flash RAM that keeps totals of the amounts being printed by the cash register program. Every single line printed that represents money is added to the appropriate accumulators. When the tax collector shows up at a store, he has the printer dump the accumulators so he knows how much the merchant sold and how much sales tax he collected.
It's a clever approach. You can try cooking the books all you want, but in the end the official receipts have to come through the register printer, and that's what the tax man reads.
Actually, taking the power away is the only historically effective solution. It will force the agencies impacted to re-evaluate their approach to surveillance, and to find a way to do it that complies with the law. That's what happened with wiretapping, and it ultimately led to the creation of the FISA court (the overseeing body you noted is missing from the current equation.)
However, the USA PATRIOT Act's current expansion of monitoring explicitly sidestepped the FISA mechanism for reasons that were never made clear. FISA was never a barrier to legitimate investigations as they approved over 99% of all interceptions, and the law was written to provide retroactive reporting to the court in the case of current activity. (Some agencies misinterpreted the clause to mean they couldn't communicate with each other, but this mistake was never tested in a courtroom -- they just hamstrung themselves under the misunderstanding.) While I'm not a fan of secret courts in a free country, I understand the need for secrecy in these cases, and FISA was a very Solomonic compromise.
Anyway, if we take it away and the FBI can't find a way to get the job done legally, then it simply was too much power in the first place.
I still don't understand why, after their failure, they didn't just stick the mint-on-a-stick straight into the open bottle? If it had reacted violently, they could have just removed it.
Or was it that they only wanted to see a blob of free-floating diet Coke explode suddenly with a mint stuck in it?
A friend is a private pilot and used to have access to a Cessna 150 Aerobat. He took me up and we went into a couple of zero-G arcs. It's astoundingly cool! And in a little Cessna it was far less than a hundred dollars an hour to play around in.
Of course, this does have its drawbacks compared to the Vomit Comet. Being a tiny(!) plane, there's no space for a passenger to actually float around the cabin. I unbuckled the seat belt so I was lifted off the seat for a while. A few objects in the cabin floated around a bit. But the little Cessna cannot achieve the speeds and altitudes required to follow a zero G parabola for more than about ten seconds at a time.
Even if it could, there's a bigger problem. Fuel intake is the limiting factor. Regular planes have a rigid fuel intake inside the gas tanks near the bottom, and the fuel sits on the bottom of the tank. The Aerobat uses "clunk tanks" similar to model planes - weighted flexible hoses in the gas tanks to ensure the fuel and intake hose are on the "bottom" of the tanks even when the plane is inverted. Both types of tanks rely on gravity to keep the fuel and the intake together. Without gravity, neither the fuel nor the intake hose are under any physical obligations to meet up with each other, and the engine can run dry. That's generally considered a "bad thing."
I always knew that cell phones are vulnerable, but to know there is a device which can basically clone your data out, with NO trace, that's downright scary! Even when LOCKED? We should start reading our contracts and our EULAs on our phone, somehow, somewhere, there's got to be something to rely on legally, if this can happen.
Such a device is called a "computer", and many people already own one. By means of a secondary device, called a "USB cable", one can attach a computer to a cell phone and read the contents from it.
If you read the "instruction manual" that comes with your cell phone, you can see plainly that a cable can be connected between the phone and the computer and the contents read from it. No phone manual I have ever read says anything about authentication of the USB cable connection. Therefore you have already been informed of as much as you need to know, legally.
The question of how much money electronics manufacturers make by screwing their customers on the cost of proprietary chargers is still left open, but you'll have to excuse me if I can't work up too much sympathy for them.
I don't think that's the case at all. The gadget manufacturers do not have any incentives to maintain the proprietary connectors. Think about it: today, no manufacturers make money selling replacement wall warts. If a customer needs to replace a wall wart, they know they can just get a universal thing at Radio Shack.
Wall warts are expensive. First, gadget manufacturers typically don't make their own, they buy them from OEMs. They add weight and bulk to the packages, which adds to shipping costs. They have to buy different wall warts for different countries with different electrical plugs and voltages, so instead of a single cheap order for 80,000, they have to place eight more expensive orders for 10,000 each. Then they have to include them in the packaging, which means carrying eight different packages and labels for eight different countries. All those differences add up to a lot of costs that are pure overhead, expenses that they wouldn't have if there was a standard that let them ship a single product.
Wall warts are a nuisance for everyone, but until recently no standards ever emerged, so no manufacturers ever had one to follow. But now there are many consumer products moving to the mini-USB connector, and I expect an avalanche of manufacturers to convert their future product lines to USB in hopes that someday they won't even have to include a charger. If they could count on their customers already having a charger in their houses, they could really save a lot of money by not including them at all.
Three phones for the Business Kings, under the Skyscrapers,
Seven for the Dwarf Tossers, in their halls of Filk,
Nine for the mortal family members, doomed to whine about overage charges,
One for the Dark Lord, sending text messages from the middle stall at work,
In the land of America, where the cellphone salesmen lie.
One ringtone to annoy them all, one ringtone to bind them,
One ringtone to identify them all and in the shopping mall find them,
In the land of America, where the cellphone salesmen lie.
I've made several of those over the years. One day I got sick of the mess and of the ugly, and so I cleaned out the entire underside of the desk of every piece of gear and strand of cable. I bought a pre-made power strip that is actually six duplex outlets side by side, spaced slightly farther apart than standard quad outlet spacing so it accommodates a wide variety of wall warts. The box itself is very solid, made of folded steel construction and is painted an old-school mottled blue enamel. It has wings with keyholes for screw mountings, and I have it screwed high up to the underside of the computer desk. I also have an older plastic Radio Shack 10-outlet surge suppressor strip that is screwed to the far back wall of the underside. I bought a cheap $10 clip-on desk lamp and have it stationed permanently on the back wall as well. Finally, I bought a couple rolls of black Velcro cable ties, and was not shy about using them. So I now have lots of power outlets (almost all of which are now full of warts anyway,) nicely bound wire bundles, and I'm not fumbling with a half-dead flashlight every time I have a new cable to plug in.
It's been a few years, and some of the wire bundles have been undone and not redone, so it's almost time to go back down there and clean them up. But I'd rather read slashdot, soo...:-)
Have you measured the voltage present at your wall jacks? 100 feet of typical Cat-3 (24 AWG) is going to add over two ohms of resistance, which might be a significant drop on your typical 5VDC connection.
I'm curious because someone cautioned me against trying to do the same thing, claiming the voltage drop would be so high it'd never work. But I never ran the math, and I never bothered to test it on a simple coil of wire. Now that I think more about it, I think the guy was full of shit. Two ohms just isn't going to affect it enough.
Well, thank you for letting me know what the payback of that effort was for you. I've decided that 13 watts here or there no longer matters in my house. My wife grows orchids under high pressure sodium lights (yes, they're the yellow street light variety), and recently purchased an orchidarium that is illuminated by some ultrabright fluorescents.
It's the coolest summer in 10 years around here, but my electric bill's shot up by an extra hundred dollars a month. As a bonus, my basement windows pour out this weird golden glow like some mutating sci-fi radiation is present. Not to mention they're shining on all these really odd plants that look like they're locked up in a giant glass case for our protection.
I could hunt down and kill every wall wart in my house, and might save two dollar's worth of electricity over the month. It's not going to change the bill by a percent.
But seriously, if $600 is out of your budget, you need to simply move the hot air out. Buy some ducting and install like a 12" fan in the ceiling (where the hot air goes) that blows down the duct and into the hallway or the closest large room. That's kinda what I do in my room at home cuz it goes up about 5 degrees Fahrenheit per hour of 3D gaming. So I blast it all out into the living room with a fan and it cools in back down to normal in about 3 minutes
That's fine for your bedroom because equipment failure means you can't frag your buddies for a couple days while you rebuild your burned-out gaming rig. But if you're running a business it's likely your business will suffer if you lose your computers. Depending on the nature of your work, a small business can maybe survive a week or two without their computers, but it'll hurt their ability to schedule work, pay the bills on time, and get money from their customers. And the longer they let it go, the more customers they'll inconvenience, and the more unpaid vendors will refuse to serve them in the future. It takes a surprisingly short amount of time before the chances are they won't recover from the outage.
The guy asked for a cost-effective solution to cooling his server room, not for cost-effective cooling during a LAN party. A cost-effective solution for a business really starts with "don't do something so stupidly cheap that you risk going out of business."
You don't even have to wait for it to fail. Telling the boss that they are in the process of failing and handing him a printout of what it costs to replace stuff, and then say "it's ok, but we're gonna need to spend a little money on A/C or that will happen." You'd be amazed at how fast wheels get greased.
I doubt it. The guy seems to think $600 is a lot of money, and that it'll somehow be "cost effective" to cobble in a window air conditioner. To me that says "update your resume, Junior, because if your boss is complaining about $600 your next paycheck is going to bounce faster than a speeding bullet."
Slashdot Mirror
How does that stop someone reading the private key from the RFID device using non-RF methods?
You mean like Ross Anderson's many successful attacks on smart cards, using microscopes, logic probes, power-fault attacks, and other really clever methods?
I'm guessing that'll be prevented mostly by 'lawyers'. "Clone this chip and we'll sue your asses, prosecute you under the DMCA, tell the RIAA you've been downloading Metallica, and unplug your freezer until your steaks thaw and make a mess all over your frozen waffles."
Now that you mention it, I've actually done that. I discovered (the hard way) that mineral deposits built up inside the valve get dislodged when it's triggered, and sometimes those chunks then get in the way of sealing the valve again. Which meant a stream of hot water that couldn't be shut off until I shut off the house valve and couldn't be turned on again until I bought a new pressure relief valve.
So now I don't do that anymore.
...will still pale in comparison to ILM. LOTR had some of the worst compositing I had ever seen. The fact that ROTK won an Oscar for Visual Effects only further illustrates the irrelevance of the politics behind the Oscars.
I don't know, I don't think ILM did a great job with Naboo or Tattooine in episodes 1-3, either. The battle of the Gungans vs the robots was a giant load of CGI all around. And that was released in 2005, while ROTK was released in 2003.
But yeah, Gollum was too aggressive for Weta, at least when they filmed it. And although the character was certainly more repulsive, Jar-jar was better done than Gollum (slightly.)
Maybe the new render farm will let them try some better algorithms or more complex models, rather than just speed up the process. We can hope.
I'm of a mixed mindset when it comes to water-cooled datacenters.
On one hand, you've got the makings of a biblical scale disaster with all that water and electricity mixing.
We had such a disaster many years ago.
The coolant lines themselves circled the walls near the ceiling of the dinosaur pen. Beneath each line was a drip tray that was alarmed to sound in case moisture was detected. These drip trays ran the entire length of the coolant lines. That is, the entire length except for about six inches in the very corner, where it was too hard for the tray installers to get to because they were behind the conduit leading down to the main power transformer. I can only assume that all that conduit also made it difficult for the plumbers to properly solder the elbow in the cooling line as well.
Of course there was only one place for the pipe to leak. The version of the tale I heard implies an impressive display of fireworks was seen by all present as the coolant entered the transformer.
Fortunately, the loss of power did not damage the mainframes (except for their ability to run.) The rooftop generator was fired up, and in short order the mainframes were back on line.
And in short order the generator engines stopped because the fuel tanks were kept almost empty, the plan being to fully fuel them only in case of need.
Once the replacement fuel was delivered, the generators ran for only a few minutes before dying again, this time for good. The ancient fuel had congealed in the injectors once the engines had stopped running.
We ran our data center for two weeks powered from a truck-based generator parked in the street while the electricians replaced the transformers and repaired the generators.
I now hear the diesel generator being started every month or two, and run for a few hours.
Rather than a biblical disaster, I'd say it was more like a Marx Brothers' movie.
Unfortunately, the only serious open-source phone project out there right now is the OpenMoko. I say unfortunately because it's a touch screen.
I don't know if your link to "Heaves-Above" was a typo, Freudian slip, or an intentional joke, but I like it a lot!
Ross Anderson at Cambridge has done some great work on the topic. There are many attacks that are documented, and I'm sure there are others. He's used a microprobing workstation to directly access chip features. They've used optical microscopes to view the state of fusible memory cells (such as you might find in a PROM.) This lets them directly read the contents of a chip. To expose the chip, they typically grind through the outer package, or use fuming nitric acid to dissolve the case.
Other, more clever attacks include active testing, where they attempt to get a functioning chip to reveal secrets while it's running. The general idea with this is that you set up a test bench that repeatedly uses the chip to encrypt a known input, change something on the chip (called "fuzzing"), and then check the output is always the same. They've used X-rays to fuzz memory locations. The X-ray beam is used to alter a memory location on the chip -- it almost doesn't matter what spot, really, as long as it causes an error in the output. You then move the X-ray beam to another spot on the chip, and repeat. The idea is that corruptions in a block cypher based on the principles of key rounds (such as DES, AES, Blowfish, Twofish, etc.) will manifest themselves as bits of key material if they corrupt the right round at the right time.
This same sort of principle was used by dropping the chip power at the right time in the encryption algorithm. By causing an error during the encryption process, that error can identify bits of the key.
It's up to you to decide if operating a micro probing workstation, building a precision guided x-ray machine, or bathing chips in fuming nitric acid is "difficult". As far as I'm concerned, yes, I think it's difficult, but I don't have access to that kind of equipment. You might have all of that down the hallway, for all I know.
Hardware advances mostly for games, media and business needs.
Actually, AC, while I don't think you are absolutely correct, you may be on to something there. It's widely believed that many of the advances in home electronics, home theater, computing and networking were due to porn more than any other factor. So if we use that as a starting base, perhaps Chrome was created for a different reason. Maybe it is really destined to be the Ultimate Porn Surfing Engine.
Just think: It'll start small. Google will use Chrome compatibility to partner with porn web operators to offer to protect their site content, and securely ensure payments. It'll work great, and soon all porn sites jump ship and start relying on the Chrome browser. Porn will no longer be viewable on IE or FF, so the world switches completely to Chrome. At that point, Google knows they have the entire internet by its collective short and curlies (almost literally), and that's when they SQUEEZE.
God help us all, we've uncovered it: Chrome's really a plot to take over the world!
Is a Google satellite evil or not evil? Discuss.
Wouldn't matter. If he gets married, it's the same thing only slower.
Screw that. Zerba.nl is nothing but flash. Do the web a favor and never sell to them.
Same thing.
It's been real for twenty years or more. IBM has long made a special cash register printer called the Fiscal Printer to meet with Italian sales tax regulations. (Other countries have since adopted the Fiscal Printer standards, but I think Italy was the first.) You can read the programming guide here: ftp://ftp.software.ibm.com/software/retail/pubs/hw/4610/3station/fiscal/italy/fit90n16.pdf
At its heart the Fiscal Printer has flash RAM that keeps totals of the amounts being printed by the cash register program. Every single line printed that represents money is added to the appropriate accumulators. When the tax collector shows up at a store, he has the printer dump the accumulators so he knows how much the merchant sold and how much sales tax he collected.
It's a clever approach. You can try cooking the books all you want, but in the end the official receipts have to come through the register printer, and that's what the tax man reads.
Actually, taking the power away is the only historically effective solution. It will force the agencies impacted to re-evaluate their approach to surveillance, and to find a way to do it that complies with the law. That's what happened with wiretapping, and it ultimately led to the creation of the FISA court (the overseeing body you noted is missing from the current equation.)
However, the USA PATRIOT Act's current expansion of monitoring explicitly sidestepped the FISA mechanism for reasons that were never made clear. FISA was never a barrier to legitimate investigations as they approved over 99% of all interceptions, and the law was written to provide retroactive reporting to the court in the case of current activity. (Some agencies misinterpreted the clause to mean they couldn't communicate with each other, but this mistake was never tested in a courtroom -- they just hamstrung themselves under the misunderstanding.) While I'm not a fan of secret courts in a free country, I understand the need for secrecy in these cases, and FISA was a very Solomonic compromise.
Anyway, if we take it away and the FBI can't find a way to get the job done legally, then it simply was too much power in the first place.
I still don't understand why, after their failure, they didn't just stick the mint-on-a-stick straight into the open bottle? If it had reacted violently, they could have just removed it.
Or was it that they only wanted to see a blob of free-floating diet Coke explode suddenly with a mint stuck in it?
A friend is a private pilot and used to have access to a Cessna 150 Aerobat. He took me up and we went into a couple of zero-G arcs. It's astoundingly cool! And in a little Cessna it was far less than a hundred dollars an hour to play around in.
Of course, this does have its drawbacks compared to the Vomit Comet. Being a tiny(!) plane, there's no space for a passenger to actually float around the cabin. I unbuckled the seat belt so I was lifted off the seat for a while. A few objects in the cabin floated around a bit. But the little Cessna cannot achieve the speeds and altitudes required to follow a zero G parabola for more than about ten seconds at a time.
Even if it could, there's a bigger problem. Fuel intake is the limiting factor. Regular planes have a rigid fuel intake inside the gas tanks near the bottom, and the fuel sits on the bottom of the tank. The Aerobat uses "clunk tanks" similar to model planes - weighted flexible hoses in the gas tanks to ensure the fuel and intake hose are on the "bottom" of the tanks even when the plane is inverted. Both types of tanks rely on gravity to keep the fuel and the intake together. Without gravity, neither the fuel nor the intake hose are under any physical obligations to meet up with each other, and the engine can run dry. That's generally considered a "bad thing."
I always knew that cell phones are vulnerable, but to know there is a device which can basically clone your data out, with NO trace, that's downright scary! Even when LOCKED? We should start reading our contracts and our EULAs on our phone, somehow, somewhere, there's got to be something to rely on legally, if this can happen.
Such a device is called a "computer", and many people already own one. By means of a secondary device, called a "USB cable", one can attach a computer to a cell phone and read the contents from it.
If you read the "instruction manual" that comes with your cell phone, you can see plainly that a cable can be connected between the phone and the computer and the contents read from it. No phone manual I have ever read says anything about authentication of the USB cable connection. Therefore you have already been informed of as much as you need to know, legally.
I don't think that's the case at all. The gadget manufacturers do not have any incentives to maintain the proprietary connectors. Think about it: today, no manufacturers make money selling replacement wall warts. If a customer needs to replace a wall wart, they know they can just get a universal thing at Radio Shack.
Wall warts are expensive. First, gadget manufacturers typically don't make their own, they buy them from OEMs. They add weight and bulk to the packages, which adds to shipping costs. They have to buy different wall warts for different countries with different electrical plugs and voltages, so instead of a single cheap order for 80,000, they have to place eight more expensive orders for 10,000 each. Then they have to include them in the packaging, which means carrying eight different packages and labels for eight different countries. All those differences add up to a lot of costs that are pure overhead, expenses that they wouldn't have if there was a standard that let them ship a single product.
Wall warts are a nuisance for everyone, but until recently no standards ever emerged, so no manufacturers ever had one to follow. But now there are many consumer products moving to the mini-USB connector, and I expect an avalanche of manufacturers to convert their future product lines to USB in hopes that someday they won't even have to include a charger. If they could count on their customers already having a charger in their houses, they could really save a lot of money by not including them at all.
Three phones for the Business Kings, under the Skyscrapers,
Seven for the Dwarf Tossers, in their halls of Filk,
Nine for the mortal family members, doomed to whine about overage charges,
One for the Dark Lord, sending text messages from the middle stall at work,
In the land of America, where the cellphone salesmen lie.
One ringtone to annoy them all, one ringtone to bind them,
One ringtone to identify them all and in the shopping mall find them,
In the land of America, where the cellphone salesmen lie.
I've made several of those over the years. One day I got sick of the mess and of the ugly, and so I cleaned out the entire underside of the desk of every piece of gear and strand of cable. I bought a pre-made power strip that is actually six duplex outlets side by side, spaced slightly farther apart than standard quad outlet spacing so it accommodates a wide variety of wall warts. The box itself is very solid, made of folded steel construction and is painted an old-school mottled blue enamel. It has wings with keyholes for screw mountings, and I have it screwed high up to the underside of the computer desk. I also have an older plastic Radio Shack 10-outlet surge suppressor strip that is screwed to the far back wall of the underside. I bought a cheap $10 clip-on desk lamp and have it stationed permanently on the back wall as well. Finally, I bought a couple rolls of black Velcro cable ties, and was not shy about using them. So I now have lots of power outlets (almost all of which are now full of warts anyway,) nicely bound wire bundles, and I'm not fumbling with a half-dead flashlight every time I have a new cable to plug in.
It's been a few years, and some of the wire bundles have been undone and not redone, so it's almost time to go back down there and clean them up. But I'd rather read slashdot, soo... :-)
Have you measured the voltage present at your wall jacks? 100 feet of typical Cat-3 (24 AWG) is going to add over two ohms of resistance, which might be a significant drop on your typical 5VDC connection.
I'm curious because someone cautioned me against trying to do the same thing, claiming the voltage drop would be so high it'd never work. But I never ran the math, and I never bothered to test it on a simple coil of wire. Now that I think more about it, I think the guy was full of shit. Two ohms just isn't going to affect it enough.
Well, thank you for letting me know what the payback of that effort was for you. I've decided that 13 watts here or there no longer matters in my house. My wife grows orchids under high pressure sodium lights (yes, they're the yellow street light variety), and recently purchased an orchidarium that is illuminated by some ultrabright fluorescents.
It's the coolest summer in 10 years around here, but my electric bill's shot up by an extra hundred dollars a month. As a bonus, my basement windows pour out this weird golden glow like some mutating sci-fi radiation is present. Not to mention they're shining on all these really odd plants that look like they're locked up in a giant glass case for our protection.
I could hunt down and kill every wall wart in my house, and might save two dollar's worth of electricity over the month. It's not going to change the bill by a percent.
It's spelled Deutsche Telekom, not Deutche.
Wow, an authentic Spelling Nazi!
But seriously, if $600 is out of your budget, you need to simply move the hot air out. Buy some ducting and install like a 12" fan in the ceiling (where the hot air goes) that blows down the duct and into the hallway or the closest large room. That's kinda what I do in my room at home cuz it goes up about 5 degrees Fahrenheit per hour of 3D gaming. So I blast it all out into the living room with a fan and it cools in back down to normal in about 3 minutes
That's fine for your bedroom because equipment failure means you can't frag your buddies for a couple days while you rebuild your burned-out gaming rig. But if you're running a business it's likely your business will suffer if you lose your computers. Depending on the nature of your work, a small business can maybe survive a week or two without their computers, but it'll hurt their ability to schedule work, pay the bills on time, and get money from their customers. And the longer they let it go, the more customers they'll inconvenience, and the more unpaid vendors will refuse to serve them in the future. It takes a surprisingly short amount of time before the chances are they won't recover from the outage.
The guy asked for a cost-effective solution to cooling his server room, not for cost-effective cooling during a LAN party. A cost-effective solution for a business really starts with "don't do something so stupidly cheap that you risk going out of business."
You don't even have to wait for it to fail. Telling the boss that they are in the process of failing and handing him a printout of what it costs to replace stuff, and then say "it's ok, but we're gonna need to spend a little money on A/C or that will happen." You'd be amazed at how fast wheels get greased.
I doubt it. The guy seems to think $600 is a lot of money, and that it'll somehow be "cost effective" to cobble in a window air conditioner. To me that says "update your resume, Junior, because if your boss is complaining about $600 your next paycheck is going to bounce faster than a speeding bullet."