What makes you think that extremely complex piece of software is supposed to be easy to setup, by just about everyone?
What makes you think everyone needs an extremely complex piece of software when their needs are often quite simple?
Apache is big and complex, nginx is small and simpler. Often one works better than the other for a particular person's needs. If someone finds Apache difficult to set up and finds nginx to be easier, then telling him to get someone to set up Apache is not the answer since he already already has the answer he is looking for. If Apache wants to be that answer too, then they need to find a way to simplify configuration.
Unity/Gnome debate is silly and you're really missing the point.
I don't think he's missing the point - Canonical pushed the small touch-screen friendly Unity on everyone, and now that they have Ubuntu running on a small formfactor touch screen that is supposed to be exactly what Unity is good at, what do they do? They dump Ubuntu entirely on that small screen and only run Ubuntu on the big monitor with no touch screen.
So tell me again what the point of Unity is if it's not for touchscreen devices?
Laser shots would be preceded by a barrage of Gummi Bears to stick to your shields and absorb the laser energy. Haribo's stock will soar due to military Gummi Bear purchases and they will become the largest and most wealthy company in the universe. Even larger than MicrAppleSoft who supplies the PADD handhelds for the fleet.
Don't talk fucking shit, for it to work you'd need to have the device on every car. The cost would be huge.
As for the control rooms & staff, they're there anyway to watch for accidents & crime.
As they said in the article, only cars that want to participate would need a transponder and that the system works well even if only few cars have them.
In my city, most people already have a transponder to pay tolls. Mine cost me $25 at Costco, and they gave me a $25 toll credit, so effective cost was zero.
They also use these same transponders to track traffic flows, using the data to provide estimated travel times to various destinations... i.e. they put up traffic signs along the freeway in Berkeley showing the expected travel time to downtown San Francisco and to SFO airport.
Plus most people have smart phones which could also become a transponder. Or GPS manufacturers could build them in to GPS's.
Such a system would add little cost to an urban smart traffic light systemduring scheduled upgrades (each traffic light costs around $100K to install) - adding a few transponder receivers and control boxes wouldn't add much to the overall cost, and only those that want to participate would need to purchase a transponder for their car. This system could replace the existing signal preemption systems set up specifically for emergency vehicles and would likely work better by letter the emergency vehicle signal its route ahead far enough to ensure all green lights. If I thought it would speed my trip through the city, I'd gladly pay $25 for a transponder or even buy a $250 transponder enabled GPS unit.
I dropped by to voice the same concern but you beat me to it. What indeed happens when one of the other monitored cars has an impulsive driver who decides in the last few seconds before the intersection to floor the accelerator?
Sorry... FAIL. This system will ONLY work if we remove humans as a variable in the equation.
How is is any different than what happens now? Right now, when when a driver faces a red light and decides to floor the accelerator, hopefully he'll get a redlight ticket, but if another car is legally in the intersection he'll t-bone that driver. Many cities already have synchronized lights to give drivers following the speed limit a green light by the time they get to the next intersection. This system just changes the synchronization to individual cars.
If drivers obey the red lights (just as they are required to now), the system doesn't fall apart if a driver changes his mind at the last minute and turns into a parking lot before he reaches the intersection - worst case, his direction will have a "wasted" green cycle, and some opposing traffic might get a red they wouldn't have otherwise had.
Seems like police would be able to broadcast a "pull over" signal that the car would obey. Or maybe the police can even signal "Go to the nearest police station" to pick up your ticket.
Seems to me that anything that it could broadcast to a car could also be replayed to any other car. There is probably something that could be done to prevent that but seems hard to do.
Right now all it takes to get your wife to pull over on a dark road is a couple of red and blue flashing lights on any type of car, so it's not like there's a whole lot of authentication going on with the present system.
The smart car will be able to validate the digital signature on the command to pull over, and further, it can contact the police department computer to validate that is a real request. Or, maybe when the cop sends a "pull over command", instead of the command going to your car, it goes to the police dispatch center, a dispatcher validates it and sends it to Ford's auto-car network, and Ford's network tells your car to pull over.
Much harder to spoof than just buying a police light-bar on the internet.
If we can make a small difference through educating Americans about working conditions in the factories where our stuff is made, I think it's an effort well worth pursuing.
But when manufacturing costs rise in China, that just means that the jobs will move to a country where workers cheaper and the Chinese factory workers will move from crappy factory jobs to no jobs at all. Which is worse?
Though I guess I can't think of any other country where wages are cheaper while still having the infrastructure to support factories.
There's little of substance to the story. It was posted so that Apple haters could rush in, out of breath, and bash Apple for "slave labor" and other goofy crap. Slashdot is a Google/Linux advocacy site, and Apple is one of their competitors.
I don't get how this is an article for Apple haters? An independent organization said that workers at factories that Appler uses have better working conditions than other factories and that the worst problem employees there face is boredom. How is that anti-Apple? I'm no Apple fan-boy, but the article validated what I thought all along - that working conditions in Apple's factories are no worse and probably better than in other factories.
Good question - maybe part of a construction permit will be entering the location and lane configuration of construction zones.
First the road is diverted to a detour while they lay a conduit across, then they re-pave this side diverting traffic to the other, then they pave the other side and traffic flips over... and this all happens in a single 8hr shift.
So for this case, the construction permit can say "Variable conditions - use manual mode".
Or maybe instead of flags and warning signs, they'll use RFID tags and RF beacons that instruct the cars what to do
a) And the national? international standards? de factor standards? for this stuff is nowhere in sight... at least not yet.
That's ok, these truly autonomous cars are nowhere in sight either, it will take time to develop standards *and* the cars. International standards aren't strictly required, there are lots of things about cars that are customized per country - driver's side of the car, headlights, crash safety regulations, emission controls etc. Automatic car signaling standards could be yet another one of those things that makes it hard to import a car.
b) Will every flag girl have all new rfid equipped stuff anytime soon? I don't think so.
Why do you keep talking as if these automatic cars are coming "soon"? There will be lots of time to sort out those things in the coming decades, and would only be neccessary on roads that are certified for automatic cars.
c) And can you imagine the security risks... anarchist types diverting highway traffic into residential areas for the lulz. Car jackers redirecting cars into dark unlit alleys...
Of course, with cars that are always connected to the internet, it's easier to validate that the beacon they are receiving is a Government Approved beacon. Fixed beacons like roadside markers can be tagged by location in a huge database kept onboard the car - if the car sees a beacon that is not where it's supposed to be, it ignores that beacon.
GPS's are getting quite good with speed limit databases -- a state that allows driverless cars should also require electronic publishing of speed limit zones. A car that detects itself going over the speed limit should refer itself for maintenance.
What could possibly go wrong? Jokers broadcasting limits of 5mph on an interstate... or 150mph in a school zone...or overriding gps signals...
That's what encryption and online validation is supposed to solve. If a construction beacon says "Speed limit 5mph", but when the car looks it up online, the database says it's really a 25mph beacon that's supposed to be across town, the car can report it and ignore it. They'd have to hack both the physical beacon and the online beacon database to reroute traffic successfully. And the car will still have onboard collision detection and road-edge and terrain detection built in plus GPS navigation so it's not like a hacker can force traffic off a cliff.
I think you're assuming driverless rather than "automatic" -- an automatic vehicle can allow driver overrides for unusual situations.
If he's going to sit there alertly paying attention for unusual situations he might as well be driving. At least it'll give him something relevant to do while he sits there. Otherwise, if the car usually gets him from his garage to the office parking lot without incident day after day, then the day the car is going to need an assist then I guarantee the driver will be watching TV, reading a book, or maybe even taking a nap in the backseat...
When they reach the level of automation that allows the car to drive itself, the driver can read a book or sleep or whatever. When the car needs help, it can say "Michael, there seems to be an unusual condition ahead, you didn't respond to my warning chime so I'm going t
But I'm curious about navigating construction sites with bizarre merges, lanes that don't line up in the slightest to the pavement markings, detours through the oncoming lane, with the crew directing traffic with hand signals...
Good question - maybe part of a construction permit will be entering the location and lane configuration of construction zones. Or maybe instead of flags and warning signs, they'll use RFID tags and RF beacons that instruct the cars what to do
I'm also curious how it copes when the car in front of it breaks down in a single lane no passing zone. Does it just sit there until a tow truck takes the car away? Does it get out of the tow trucks way if necessary.
I always assumed the cars would have a manual override - after all, it doesn't know what parking space I want, it might want to park me in the spot closest to my destination, but I may want to park under the tree in the shade (or *not* want to park under the tree so the birds on the tree don't poop on my car) so I'll want to be able to guide the car to do what I want.
How it deals with a police car trying to pull it over for a broken tail light.
Seems like police would be able to broadcast a "pull over" signal that the car would obey. Or maybe the police can even signal "Go to the nearest police station" to pick up your ticket.
How does it know the speed limit at every segment of road, and if the limits change how does it get updated... what happens if it blows through a speed trap a 1 mph over the limit.
GPS's are getting quite good with speed limit databases -- a state that allows driverless cars should also require electronic publishing of speed limit zones. A car that detects itself going over the speed limit should refer itself for maintenance. If someone modifies a car to break the speed limit, they should be fined appropriately. Car manufacturers should be responsible if a non-tampered car exceeds the speed limit.
Does it obey the officer trying to make eye contact with the nonexistent driver while he holds his hand up to wave you down.
How does it know a police officer is waving it down and not some kids... will it drive into a back alley or field everytime someone steps onto the road and points at one?
Its one thing to leave enough distance in front, and to safely come to a stop if the car in front of you has an accident.
But then what? Can it move out of the way of the rescue vehicles? Can it navigate the makeshift detour the police are directing?
I think you're assuming driverless rather than "automatic" -- an automatic vehicle can allow driver overrides for unusual situations.
I think we are decades away from practical driverless vehicles that could operate without a qualified driver sitting there ready to take over. At best we'll need to see a driver in the seat at all times, ready to take over when the car reports that its unable to proceed. Hopefully the car can at least detect situations it can't manage well enough... but im skeptical of it even being able to do that.
Oh, I agree that 100% automatic cars are decades away. For now, they will be driver assisting vehicles.
I always figured that there will be special roads that are certified for automatic operation (and *only* for automatic cars - no manual drivers on those roads) -- those roads would follow guidelines for electronically readable signage, roadside markers, etc. When off of those roads, then the car would switch to driver-assisted operation.
I've seen lots of video of them under ideal scenarios.
Let's get some crash video!:)
Hydroplaning, black ice, big potholes, road debris, silver-hairs stomping on the brakes, et cetera.
Should be entertaining, if disconcerting, to say the least.
I would guess that automatic cars would be even better than humans at any of these conditions:
Hydroplaning, black ice: The car has an excellent sense of traction of each wheel and can take the appropriate action (like reducing power to the slipping wheel(s)) without overreacting (slamming on the brakes because the driver feels the car start to slide), and can take proactive measures when sensors detect hazardous conditions (i.e. drive slower when ice/snow/heavy rain are detected rather than thinking "Hey, it's an SUV, I can drive 75mph through the snow, just like on the commercials.). Depending on what kind of sensors they use, it might be able to see perfectly through heavy fog/rain and in the dark.
Big potholes: I've never hit a pothole big enough to make me feel like I was going to lose control, but since the automatic car always has a firm grip on the wheel and is never distracted by changing radio stations or talking on the phone, I'd say it can do a better job.
Silver-hairs stomping on the brakes? It's the young drivers in sports cars that seem to have that problem -- most older drivers seem to drive slow and brake way before they need to. But this is where the automatic car has the real advantage - not only will it keep a safe following distance, but its sensors will sense the stopping car ahead and calculate the closing rate and will know exactly when it needs to panic stop, and it can do it faster than a human (which for most people means around 250ms just to recognize the threat and then a couple hundred ms to get pressure on the brakes). And the car will sense it regardless of whether or not the brake lights are working on the car ahead of you, and will stop even if it's not a car. If the sensor pod is on top of the car, that extra foot or so of height will help it see over hill crests better than a driver.
I hit a refrigerator on the road one dark night, an automatic car probably would have avoided it completely. I saw a pickup backing up on the road shoulder and was focused on him to see what he was up to and didn't see the refrigerator that had fallen off his truck until it was too late to stop. I started to slow when I saw the truck and hit the brakes as soon as I did see the refrigerator so was only going around 25mph when I hit it. The 'fridge was totaled, but the only damage to my car was a broken front turn signal.
Granted, a heads up display linked to sensors might be able to feed all of the same information to a driver that an automatic car has (threat ahead - object in road! Losing traction in left-front and left-rear wheels! You will have a collision, brake now and turn wheel 10 degrees to the left!) But it's still relying on the driver to react correctly and react quickly.
When it happened to me, I burnt some DVD's and gave her a copy of everything that was hers and of everything that was "ours" (shared pictures, etc).
We didn't have a shared domain, but if you can't decide on an equitable solution, then just keep it as a shared domain. She gets the password with admin rights, you get the password with admin rights. Get your own separate domains and migrate things there. The only shared account we had was Netflix and we canceled that and each started our own account.
The dog was the hardest to "split" - we ended up with a shared custody arrangement that was actually quite convenient for both of us - we got part-time dog ownership and a built-in dogsitter for vacations. Worked out well for a few years until she moved away to another state. She moved to a nice house + yard, so she got the dog.
k. This is slightly laughable. 33,500 rows? in 87 seconds? that seems glacial. And 1.23 seconds being the new speed? that seems as expected.
aside from comparing the speed of a Lada vs a common garden slug, how does this compare against other databases?
It was actually only 2000 rows returned -- the source tables had 35K rows. But in fairness, this was for a join across 11 tables, and it's easy for a database to perform slowly at this task if it doesn't have a good query optimizer. A badly optimized query could end up doing thousands (or even millions) of full table scans of multi-thousand row tables.
As long as the manufacturer can document that the quantity of legitimate drugs shipped match their use of amphetamine salts, what is the point of quotas? Why is the DEA involved at all in projecting and limiting the amount of any prescription drug?
If there's a problem with amphetamine salt derived drugs being overprescribed, that sounds like a problem for the FDA, not the DEA.
If there's a problem with the theft of amphetamine derived drugs, that problem will exist regardless of any quotas.
If there's a problem with theft of the amphetamine salts from manufacturers storerooms, then set high fines for loss of any material that's not accounted for in sales of legitimate drugs, making sure they will use appropriate security.
What benefit is there for restricting the amphetamine salts sales to drug manufacturers other than to drive up prices by causing false scarcity?
Dont use data! Wifi only people.. if we all do so together for a week, they wil be forced to start charging less for something that costs them less to use in the first place.
yes your texts and FB posts all use less data than a simple 2min phone call. so why are you paying for it?
I fail to see why AT&T would care if everyone stopped using data for a week? It's not going to reduce their revenue, everyone has already paid for their data use for the month. All it will do is reduce their network congestion for a week and people that choose not to participate will experience better service. It's not like I can use voice to replace my data usage - I'm not going to call home and ask my wife to prop her phone up next to the stereo as a Pandora substitute, nor am I going to call all of my friends to tell them that I just saw the most amazing sunrise as a substitute for posting it on my facebook wall so they can ignore it at their leisure.
What lesson is this supposed to teach AT&T?
This sounds like the misguided "protests" against gas prices by telling everyone to not buy gas on Tuesday. All it does is makes everyone shift demand to another day. Now if you could convince people to give up driving for a month (or to cancel their data plan for a month (which of course, they can't do under most cellular contracts)), that might send a real message.
I think the best tactic would be to sue AT&T for breach of contract by selling an "unlimited" service and then capping that service lower than 3GB and 5GB "limited" plans. I could understand if they rate-limited only during peak periods, but why 24 hours a day until the end of the billing cycle? If they can't support unlimited plans, they stop selling them and stop grandfathering people in when they renew their contract - they have the right to set new terms at the end of the contract, so they should just stop selling unlimited plans they have no intention of fulfilling. (of course, I know the answer - they grandfather "unlimited" plans because that's the only carrot they can dangle to get someone to stay with AT&T when their contract is up)
Nuclear power plants tend to be much larger so they have a lot more waste heat to dump. In addition, some forms of fossil fuel plants dump their waste heat directly into the air without using water cooling. This works because the combustion temperature inside a fossil-fuel power plant is much higher than the fuel plate temperatures in a water-cooled nuclear power plant so they can still be efficient overall even with using the atmosphere as a heat sink.
One interesting article I read said that power generation accounts for about half the water usage in the USA:
I think few people use the SSL keys on the home router/firewall for anything other than local administration of their firewall so it doesn't really matter if the SSL can be broken, no one is watching and no one cares.
Even if the few people that actually used their home router/firewall to encrypt data sent over the public internet have hackable encrypted sessions, it's pretty unlikely that an attacker is watching their packets on the off chance that they have a hackable key when there are far easier and more common vulnerabilities to exploit when the attacker has access to your network packets (like firesheep style session stealing).
Picture quality is pretty far down my list of things that make me watch my TV. I watch a lot of Netflix streaming, and though I'm amazed at the jump in quality when I'm watching a Bluray, even when I own the Bluray I'm more likely to just pull it up on Netflix when i want to watch it again than dig through my movie collection to find it and load it into the bluray player. The quality of the picture really does nothing for me.
The fact that it takes minutes for the disk to load, then more minutes to skip through the previews doesn't help. Why couldn't Bluray have just been a better quality DVD? Why do I have to wait so long for a disk to start up and why in gods name do I need a firmware update before some disks will play!?
Here's one thing that *would* make me buy a new TV: put in a separate closed caption display just under (or above) the main monitor and let me view the captions on that display. There are many foreign films that I love, but I hate that some of the picture is obscured by the captions.
Pretty sure this is the debian bug where the relevant packages for whatever reason decided to use a static number for the rng seed. There is an xkcd which pokes fun at this.
Oh good, it's not too late to post the obligatory XKCD link:
Much like I don't bypass my electric meter even though I know how to do it it, but I'm not willing to accept the risk or danger of doing so.
I don't do it because there is a seal on it that I would have to remove to bypass the meter, but then I will have to pay a huge fine. If the meter was not sealed or bypassing it would not be against the law I would have done it already (or at least I would reset it once in a while - less evidence than would be if I bypassed it)
So you're ok with stealing power from a hotel, but you draw the line at stealing from a large utility?
If you don't know how to bypass your meter without breaking the seal on the meter, then maybe you shouldn't be messing with live power outlets. Even the pot growers rarely break meter seals for their grow houses since that's too easily discovered.
By the time this technology is commonplace, hotels will have non-replaceable LED lighting that's built in to the fixtures (never needs replacing over the lifetime of the hotel room), and maybe the whole room's lighting will be low voltage LED lighting powered by a single controller so it all (Lights, TV, heat, cooling, etc) will all be controlled by a single remote control.
As long as there is enough power in a easily accessible wire, I can bring an adapter that would make the power there compatible with my stuff.
In my low-voltage LED lit utopia, I was envisioning no exposed wires - LED"s would be built into the ceiling or around the bathroom mirror with wiring inaccessible behind the walls, not as standalone plug-in lamps that you can use a vampire clip to steal from.
Speaking as an electrician, I cannot imagine this ever becoming widespread. The ability to control power to and from each socket has existed for years. The reason we don't use it is mostly cost (each outlet needs a dedicated line back to a relay bank or a PLC).
This new system is a little different in how it works, but cost is still going to be a huge factor involved, as well as practicallity. Most outlet boxes i've seen don't have the space to put in a GFCI let alone something as complicated as this.
But the system becomes much cheaper when you can move control to the endpoints and use existing wiring for communications. If these outlets are mass produced they could be very cheap
A hotel doesn't have to make much money per guest to make this worthwhile. If they install a $10000 controller, and pay $100/room replacing outlets in a 200 room hotel (labor can be effectively free if they do during periodic room refurbishing), that's $30,000 in capital costs. If they average 50% occupancy and earn just $1/day/room in power fees, that's 365 * 200 * 50% * $1 = $36K per year. So the system has paid for itself after the first year.
Will I really pay per minute to charge my laptop, or will I go buy a screwdriver and some $5 alligator clips?
I can't speak for what *you* will do, but I don't stick my fingers anywhere near a bare power conductor, so I'd pay for the power. Much like I don't bypass my electric meter even though I know how to do it it, but I'm not willing to accept the risk or danger of doing so.
Better yet, will I just get one of those light socket plug adapters? Either way, I'm not paying my hotel for power when I travel. Many already try to wing you for $15 or more just for 24 hours of slow internet access...
By the time this technology is commonplace, hotels will have non-replaceable LED lighting that's built in to the fixtures (never needs replacing over the lifetime of the hotel room), and maybe the whole room's lighting will be low voltage LED lighting powered by a single controller so it all (Lights, TV, heat, cooling, etc) will all be controlled by a single remote control.
and don't get me started on the minibars!
Hey - are you the guy that refilled the mini Jack Daniel's bottle with iced tea!?
Even if you could do this, why would you want to? Do you want to pay the hotel every time you use your pay-for-power enabled toaster at home?
This isn't a device to let the power company monitor your power usage, it's a way for the facility owner to do so (and possible to charge you for usage). It requires an on-premises controller to communicate with the outlets. When you bring the hotel's outlet home, it won't talk to your controller (well, probably not) and certainly won't send a bill for your usage to the hotel.
... because I'm just looking for someone else to blame, too. But there is this big WTF:
The possibility of being able to capture this data is especially probable since Denver International offers free WiFi and it is an unencrypted network.
It doesn't have to be unencrypted to be free.
Well, if you want a secure encrypted network, it's probably not going to be free.
There's only moderate additional security gained by having a WPA encrypted network where everyone has the same PSK since it's trivial to capture the association handshake (by forcing them to reassociate if neccessary) and steal the session key from anyone's session - Wireshark will do this for you. Alternatively, you can set up a hotspot on your laptop called "SouthwestAirlines" and nearby clients will connect to your laptop instead of the real Southwest network and you can capture all of their packets.
To make a secure encrypted network, they'd need to implement something like 802.1x security with unique username/passwords for each user and with Wifi clients configured to authenticate the network's 802.1x certificate (to prevent someone from setting up a rogue SouthwestAirlines access point).
Few providers of free Wifi service are going to be willing to run a helpdesk to assist all of the users with setting this up - it's not always trivial (depending on the device). So it's probably better to not provide the illusion of a secure encrypted network when it's not. The users that are sophisticated enough to set up 802.1x authentication on their device are probably also sophisticated to use a VPN to secure their data.
When I connect via an open Wifi network, I always VPN to my company or my home internet router so all of my wifi traffic is encrypted.
Slashdot Mirror
What makes you think that extremely complex piece of software is supposed to be easy to setup, by just about everyone?
What makes you think everyone needs an extremely complex piece of software when their needs are often quite simple?
Apache is big and complex, nginx is small and simpler. Often one works better than the other for a particular person's needs. If someone finds Apache difficult to set up and finds nginx to be easier, then telling him to get someone to set up Apache is not the answer since he already already has the answer he is looking for. If Apache wants to be that answer too, then they need to find a way to simplify configuration.
Unity/Gnome debate is silly and you're really missing the point.
I don't think he's missing the point - Canonical pushed the small touch-screen friendly Unity on everyone, and now that they have Ubuntu running on a small formfactor touch screen that is supposed to be exactly what Unity is good at, what do they do? They dump Ubuntu entirely on that small screen and only run Ubuntu on the big monitor with no touch screen.
So tell me again what the point of Unity is if it's not for touchscreen devices?
Laser Beams.
That's all.
my mirror shields will take the day
Laser shots would be preceded by a barrage of Gummi Bears to stick to your shields and absorb the laser energy. Haribo's stock will soar due to military Gummi Bear purchases and they will become the largest and most wealthy company in the universe. Even larger than MicrAppleSoft who supplies the PADD handhelds for the fleet.
Don't talk fucking shit, for it to work you'd need to have the device on every car. The cost would be huge.
As for the control rooms & staff, they're there anyway to watch for accidents & crime.
As they said in the article, only cars that want to participate would need a transponder and that the system works well even if only few cars have them.
In my city, most people already have a transponder to pay tolls. Mine cost me $25 at Costco, and they gave me a $25 toll credit, so effective cost was zero.
They also use these same transponders to track traffic flows, using the data to provide estimated travel times to various destinations... i.e. they put up traffic signs along the freeway in Berkeley showing the expected travel time to downtown San Francisco and to SFO airport.
Plus most people have smart phones which could also become a transponder. Or GPS manufacturers could build them in to GPS's.
Such a system would add little cost to an urban smart traffic light systemduring scheduled upgrades (each traffic light costs around $100K to install) - adding a few transponder receivers and control boxes wouldn't add much to the overall cost, and only those that want to participate would need to purchase a transponder for their car. This system could replace the existing signal preemption systems set up specifically for emergency vehicles and would likely work better by letter the emergency vehicle signal its route ahead far enough to ensure all green lights. If I thought it would speed my trip through the city, I'd gladly pay $25 for a transponder or even buy a $250 transponder enabled GPS unit.
I dropped by to voice the same concern but you beat me to it. What indeed happens when one of the other monitored cars has an impulsive driver who decides in the last few seconds before the intersection to floor the accelerator?
Sorry... FAIL. This system will ONLY work if we remove humans as a variable in the equation.
How is is any different than what happens now? Right now, when when a driver faces a red light and decides to floor the accelerator, hopefully he'll get a redlight ticket, but if another car is legally in the intersection he'll t-bone that driver. Many cities already have synchronized lights to give drivers following the speed limit a green light by the time they get to the next intersection. This system just changes the synchronization to individual cars.
If drivers obey the red lights (just as they are required to now), the system doesn't fall apart if a driver changes his mind at the last minute and turns into a parking lot before he reaches the intersection - worst case, his direction will have a "wasted" green cycle, and some opposing traffic might get a red they wouldn't have otherwise had.
Seems like police would be able to broadcast a "pull over" signal that the car would obey. Or maybe the police can even signal "Go to the nearest police station" to pick up your ticket.
Seems to me that anything that it could broadcast to a car could also be replayed to any other car. There is probably something that could be done to prevent that but seems hard to do.
Right now all it takes to get your wife to pull over on a dark road is a couple of red and blue flashing lights on any type of car, so it's not like there's a whole lot of authentication going on with the present system.
The smart car will be able to validate the digital signature on the command to pull over, and further, it can contact the police department computer to validate that is a real request. Or, maybe when the cop sends a "pull over command", instead of the command going to your car, it goes to the police dispatch center, a dispatcher validates it and sends it to Ford's auto-car network, and Ford's network tells your car to pull over.
Much harder to spoof than just buying a police light-bar on the internet.
If we can make a small difference through educating Americans about working conditions in the factories where our stuff is made, I think it's an effort well worth pursuing.
But when manufacturing costs rise in China, that just means that the jobs will move to a country where workers cheaper and the Chinese factory workers will move from crappy factory jobs to no jobs at all. Which is worse?
Though I guess I can't think of any other country where wages are cheaper while still having the infrastructure to support factories.
There's little of substance to the story. It was posted so that Apple haters could rush in, out of breath, and bash Apple for "slave labor" and other goofy crap. Slashdot is a Google/Linux advocacy site, and Apple is one of their competitors.
I don't get how this is an article for Apple haters? An independent organization said that workers at factories that Appler uses have better working conditions than other factories and that the worst problem employees there face is boredom. How is that anti-Apple? I'm no Apple fan-boy, but the article validated what I thought all along - that working conditions in Apple's factories are no worse and probably better than in other factories.
Good question - maybe part of a construction permit will be entering the location and lane configuration of construction zones.
First the road is diverted to a detour while they lay a conduit across, then they re-pave this side diverting traffic to the other, then they pave the other side and traffic flips over... and this all happens in a single 8hr shift.
So for this case, the construction permit can say "Variable conditions - use manual mode".
Or maybe instead of flags and warning signs, they'll use RFID tags and RF beacons that instruct the cars what to do
a) And the national? international standards? de factor standards? for this stuff is nowhere in sight... at least not yet.
That's ok, these truly autonomous cars are nowhere in sight either, it will take time to develop standards *and* the cars. International standards aren't strictly required, there are lots of things about cars that are customized per country - driver's side of the car, headlights, crash safety regulations, emission controls etc. Automatic car signaling standards could be yet another one of those things that makes it hard to import a car.
b) Will every flag girl have all new rfid equipped stuff anytime soon? I don't think so.
Why do you keep talking as if these automatic cars are coming "soon"? There will be lots of time to sort out those things in the coming decades, and would only be neccessary on roads that are certified for automatic cars.
c) And can you imagine the security risks ... anarchist types diverting highway traffic into residential areas for the lulz. Car jackers redirecting cars into dark unlit alleys...
Of course, with cars that are always connected to the internet, it's easier to validate that the beacon they are receiving is a Government Approved beacon. Fixed beacons like roadside markers can be tagged by location in a huge database kept onboard the car - if the car sees a beacon that is not where it's supposed to be, it ignores that beacon.
GPS's are getting quite good with speed limit databases -- a state that allows driverless cars should also require electronic publishing of speed limit zones. A car that detects itself going over the speed limit should refer itself for maintenance.
What could possibly go wrong? Jokers broadcasting limits of 5mph on an interstate... or 150mph in a school zone...or overriding gps signals...
That's what encryption and online validation is supposed to solve. If a construction beacon says "Speed limit 5mph", but when the car looks it up online, the database says it's really a 25mph beacon that's supposed to be across town, the car can report it and ignore it. They'd have to hack both the physical beacon and the online beacon database to reroute traffic successfully. And the car will still have onboard collision detection and road-edge and terrain detection built in plus GPS navigation so it's not like a hacker can force traffic off a cliff.
I think you're assuming driverless rather than "automatic" -- an automatic vehicle can allow driver overrides for unusual situations.
If he's going to sit there alertly paying attention for unusual situations he might as well be driving. At least it'll give him something relevant to do while he sits there. Otherwise, if the car usually gets him from his garage to the office parking lot without incident day after day, then the day the car is going to need an assist then I guarantee the driver will be watching TV, reading a book, or maybe even taking a nap in the backseat...
When they reach the level of automation that allows the car to drive itself, the driver can read a book or sleep or whatever. When the car needs help, it can say "Michael, there seems to be an unusual condition ahead, you didn't respond to my warning chime so I'm going t
But I'm curious about navigating construction sites with bizarre merges, lanes that don't line up in the slightest to the pavement markings, detours through the oncoming lane, with the crew directing traffic with hand signals...
Good question - maybe part of a construction permit will be entering the location and lane configuration of construction zones. Or maybe instead of flags and warning signs, they'll use RFID tags and RF beacons that instruct the cars what to do
I'm also curious how it copes when the car in front of it breaks down in a single lane no passing zone. Does it just sit there until a tow truck takes the car away? Does it get out of the tow trucks way if necessary.
I always assumed the cars would have a manual override - after all, it doesn't know what parking space I want, it might want to park me in the spot closest to my destination, but I may want to park under the tree in the shade (or *not* want to park under the tree so the birds on the tree don't poop on my car) so I'll want to be able to guide the car to do what I want.
How it deals with a police car trying to pull it over for a broken tail light.
Seems like police would be able to broadcast a "pull over" signal that the car would obey. Or maybe the police can even signal "Go to the nearest police station" to pick up your ticket.
How does it know the speed limit at every segment of road, and if the limits change how does it get updated... what happens if it blows through a speed trap a 1 mph over the limit.
GPS's are getting quite good with speed limit databases -- a state that allows driverless cars should also require electronic publishing of speed limit zones. A car that detects itself going over the speed limit should refer itself for maintenance. If someone modifies a car to break the speed limit, they should be fined appropriately. Car manufacturers should be responsible if a non-tampered car exceeds the speed limit.
Does it obey the officer trying to make eye contact with the nonexistent driver while he holds his hand up to wave you down.
How does it know a police officer is waving it down and not some kids... will it drive into a back alley or field everytime someone steps onto the road and points at one?
Its one thing to leave enough distance in front, and to safely come to a stop if the car in front of you has an accident.
But then what? Can it move out of the way of the rescue vehicles? Can it navigate the makeshift detour the police are directing?
I think you're assuming driverless rather than "automatic" -- an automatic vehicle can allow driver overrides for unusual situations.
I think we are decades away from practical driverless vehicles that could operate without a qualified driver sitting there ready to take over. At best we'll need to see a driver in the seat at all times, ready to take over when the car reports that its unable to proceed. Hopefully the car can at least detect situations it can't manage well enough... but im skeptical of it even being able to do that.
Oh, I agree that 100% automatic cars are decades away. For now, they will be driver assisting vehicles.
I always figured that there will be special roads that are certified for automatic operation (and *only* for automatic cars - no manual drivers on those roads) -- those roads would follow guidelines for electronically readable signage, roadside markers, etc. When off of those roads, then the car would switch to driver-assisted operation.
I've seen lots of video of them under ideal scenarios.
Let's get some crash video! :)
Hydroplaning, black ice, big potholes, road debris, silver-hairs stomping on the brakes, et cetera.
Should be entertaining, if disconcerting, to say the least.
I would guess that automatic cars would be even better than humans at any of these conditions:
Hydroplaning, black ice: The car has an excellent sense of traction of each wheel and can take the appropriate action (like reducing power to the slipping wheel(s)) without overreacting (slamming on the brakes because the driver feels the car start to slide), and can take proactive measures when sensors detect hazardous conditions (i.e. drive slower when ice/snow/heavy rain are detected rather than thinking "Hey, it's an SUV, I can drive 75mph through the snow, just like on the commercials.). Depending on what kind of sensors they use, it might be able to see perfectly through heavy fog/rain and in the dark.
Big potholes: I've never hit a pothole big enough to make me feel like I was going to lose control, but since the automatic car always has a firm grip on the wheel and is never distracted by changing radio stations or talking on the phone, I'd say it can do a better job.
Silver-hairs stomping on the brakes? It's the young drivers in sports cars that seem to have that problem -- most older drivers seem to drive slow and brake way before they need to. But this is where the automatic car has the real advantage - not only will it keep a safe following distance, but its sensors will sense the stopping car ahead and calculate the closing rate and will know exactly when it needs to panic stop, and it can do it faster than a human (which for most people means around 250ms just to recognize the threat and then a couple hundred ms to get pressure on the brakes). And the car will sense it regardless of whether or not the brake lights are working on the car ahead of you, and will stop even if it's not a car. If the sensor pod is on top of the car, that extra foot or so of height will help it see over hill crests better than a driver.
I hit a refrigerator on the road one dark night, an automatic car probably would have avoided it completely. I saw a pickup backing up on the road shoulder and was focused on him to see what he was up to and didn't see the refrigerator that had fallen off his truck until it was too late to stop. I started to slow when I saw the truck and hit the brakes as soon as I did see the refrigerator so was only going around 25mph when I hit it. The 'fridge was totaled, but the only damage to my car was a broken front turn signal.
Granted, a heads up display linked to sensors might be able to feed all of the same information to a driver that an automatic car has (threat ahead - object in road! Losing traction in left-front and left-rear wheels! You will have a collision, brake now and turn wheel 10 degrees to the left!) But it's still relying on the driver to react correctly and react quickly.
When it happened to me, I burnt some DVD's and gave her a copy of everything that was hers and of everything that was "ours" (shared pictures, etc).
We didn't have a shared domain, but if you can't decide on an equitable solution, then just keep it as a shared domain. She gets the password with admin rights, you get the password with admin rights. Get your own separate domains and migrate things there. The only shared account we had was Netflix and we canceled that and each started our own account.
The dog was the hardest to "split" - we ended up with a shared custody arrangement that was actually quite convenient for both of us - we got part-time dog ownership and a built-in dogsitter for vacations. Worked out well for a few years until she moved away to another state. She moved to a nice house + yard, so she got the dog.
k. This is slightly laughable. 33,500 rows? in 87 seconds? that seems glacial. And 1.23 seconds being the new speed? that seems as expected.
aside from comparing the speed of a Lada vs a common garden slug, how does this compare against other databases?
It was actually only 2000 rows returned -- the source tables had 35K rows. But in fairness, this was for a join across 11 tables, and it's easy for a database to perform slowly at this task if it doesn't have a good query optimizer. A badly optimized query could end up doing thousands (or even millions) of full table scans of multi-thousand row tables.
As long as the manufacturer can document that the quantity of legitimate drugs shipped match their use of amphetamine salts, what is the point of quotas? Why is the DEA involved at all in projecting and limiting the amount of any prescription drug?
If there's a problem with amphetamine salt derived drugs being overprescribed, that sounds like a problem for the FDA, not the DEA.
If there's a problem with the theft of amphetamine derived drugs, that problem will exist regardless of any quotas.
If there's a problem with theft of the amphetamine salts from manufacturers storerooms, then set high fines for loss of any material that's not accounted for in sales of legitimate drugs, making sure they will use appropriate security.
What benefit is there for restricting the amphetamine salts sales to drug manufacturers other than to drive up prices by causing false scarcity?
Dont use data! Wifi only people.. if we all do so together for a week, they wil be forced to start charging less for something that costs them less to use in the first place.
yes your texts and FB posts all use less data than a simple 2min phone call. so why are you paying for it?
I fail to see why AT&T would care if everyone stopped using data for a week? It's not going to reduce their revenue, everyone has already paid for their data use for the month. All it will do is reduce their network congestion for a week and people that choose not to participate will experience better service. It's not like I can use voice to replace my data usage - I'm not going to call home and ask my wife to prop her phone up next to the stereo as a Pandora substitute, nor am I going to call all of my friends to tell them that I just saw the most amazing sunrise as a substitute for posting it on my facebook wall so they can ignore it at their leisure.
What lesson is this supposed to teach AT&T?
This sounds like the misguided "protests" against gas prices by telling everyone to not buy gas on Tuesday. All it does is makes everyone shift demand to another day. Now if you could convince people to give up driving for a month (or to cancel their data plan for a month (which of course, they can't do under most cellular contracts)), that might send a real message.
I think the best tactic would be to sue AT&T for breach of contract by selling an "unlimited" service and then capping that service lower than 3GB and 5GB "limited" plans. I could understand if they rate-limited only during peak periods, but why 24 hours a day until the end of the billing cycle? If they can't support unlimited plans, they stop selling them and stop grandfathering people in when they renew their contract - they have the right to set new terms at the end of the contract, so they should just stop selling unlimited plans they have no intention of fulfilling. (of course, I know the answer - they grandfather "unlimited" plans because that's the only carrot they can dangle to get someone to stay with AT&T when their contract is up)
Nuclear power plants tend to be much larger so they have a lot more waste heat to dump. In addition, some forms of fossil fuel plants dump their waste heat directly into the air without using water cooling. This works because the combustion temperature inside a fossil-fuel power plant is much higher than the fuel plate temperatures in a water-cooled nuclear power plant so they can still be efficient overall even with using the atmosphere as a heat sink.
One interesting article I read said that power generation accounts for about half the water usage in the USA:
http://www.scientificamerican.com/article.cfm?id=how-saving-energy-means-conserving-water
Don't fossil fueled plants also have waste heat they need to dump somewhere? Do Nukes generate a lot more waste heat?
I think few people use the SSL keys on the home router/firewall for anything other than local administration of their firewall so it doesn't really matter if the SSL can be broken, no one is watching and no one cares.
Even if the few people that actually used their home router/firewall to encrypt data sent over the public internet have hackable encrypted sessions, it's pretty unlikely that an attacker is watching their packets on the off chance that they have a hackable key when there are far easier and more common vulnerabilities to exploit when the attacker has access to your network packets (like firesheep style session stealing).
Picture quality is pretty far down my list of things that make me watch my TV. I watch a lot of Netflix streaming, and though I'm amazed at the jump in quality when I'm watching a Bluray, even when I own the Bluray I'm more likely to just pull it up on Netflix when i want to watch it again than dig through my movie collection to find it and load it into the bluray player. The quality of the picture really does nothing for me.
The fact that it takes minutes for the disk to load, then more minutes to skip through the previews doesn't help. Why couldn't Bluray have just been a better quality DVD? Why do I have to wait so long for a disk to start up and why in gods name do I need a firmware update before some disks will play!?
Here's one thing that *would* make me buy a new TV: put in a separate closed caption display just under (or above) the main monitor and let me view the captions on that display. There are many foreign films that I love, but I hate that some of the picture is obscured by the captions.
Pretty sure this is the debian bug where the relevant packages for whatever reason decided to use a static number for the rng seed. There is an xkcd which pokes fun at this.
Oh good, it's not too late to post the obligatory XKCD link:
http://xkcd.com/221/
Much like I don't bypass my electric meter even though I know how to do it it, but I'm not willing to accept the risk or danger of doing so.
I don't do it because there is a seal on it that I would have to remove to bypass the meter, but then I will have to pay a huge fine. If the meter was not sealed or bypassing it would not be against the law I would have done it already (or at least I would reset it once in a while - less evidence than would be if I bypassed it)
So you're ok with stealing power from a hotel, but you draw the line at stealing from a large utility?
If you don't know how to bypass your meter without breaking the seal on the meter, then maybe you shouldn't be messing with live power outlets. Even the pot growers rarely break meter seals for their grow houses since that's too easily discovered.
By the time this technology is commonplace, hotels will have non-replaceable LED lighting that's built in to the fixtures (never needs replacing over the lifetime of the hotel room), and maybe the whole room's lighting will be low voltage LED lighting powered by a single controller so it all (Lights, TV, heat, cooling, etc) will all be controlled by a single remote control.
As long as there is enough power in a easily accessible wire, I can bring an adapter that would make the power there compatible with my stuff.
In my low-voltage LED lit utopia, I was envisioning no exposed wires - LED"s would be built into the ceiling or around the bathroom mirror with wiring inaccessible behind the walls, not as standalone plug-in lamps that you can use a vampire clip to steal from.
Speaking as an electrician, I cannot imagine this ever becoming widespread. The ability to control power to and from each socket has existed for years. The reason we don't use it is mostly cost (each outlet needs a dedicated line back to a relay bank or a PLC).
This new system is a little different in how it works, but cost is still going to be a huge factor involved, as well as practicallity. Most outlet boxes i've seen don't have the space to put in a GFCI let alone something as complicated as this.
But the system becomes much cheaper when you can move control to the endpoints and use existing wiring for communications. If these outlets are mass produced they could be very cheap
A hotel doesn't have to make much money per guest to make this worthwhile. If they install a $10000 controller, and pay $100/room replacing outlets in a 200 room hotel (labor can be effectively free if they do during periodic room refurbishing), that's $30,000 in capital costs. If they average 50% occupancy and earn just $1/day/room in power fees, that's 365 * 200 * 50% * $1 = $36K per year. So the system has paid for itself after the first year.
Will I really pay per minute to charge my laptop, or will I go buy a screwdriver and some $5 alligator clips?
I can't speak for what *you* will do, but I don't stick my fingers anywhere near a bare power conductor, so I'd pay for the power. Much like I don't bypass my electric meter even though I know how to do it it, but I'm not willing to accept the risk or danger of doing so.
Better yet, will I just get one of those light socket plug adapters? Either way, I'm not paying my hotel for power when I travel. Many already try to wing you for $15 or more just for 24 hours of slow internet access...
By the time this technology is commonplace, hotels will have non-replaceable LED lighting that's built in to the fixtures (never needs replacing over the lifetime of the hotel room), and maybe the whole room's lighting will be low voltage LED lighting powered by a single controller so it all (Lights, TV, heat, cooling, etc) will all be controlled by a single remote control.
and don't get me started on the minibars!
Hey - are you the guy that refilled the mini Jack Daniel's bottle with iced tea!?
Even if you could do this, why would you want to? Do you want to pay the hotel every time you use your pay-for-power enabled toaster at home?
This isn't a device to let the power company monitor your power usage, it's a way for the facility owner to do so (and possible to charge you for usage). It requires an on-premises controller to communicate with the outlets. When you bring the hotel's outlet home, it won't talk to your controller (well, probably not) and certainly won't send a bill for your usage to the hotel.
... because I'm just looking for someone else to blame, too. But there is this big WTF:
The possibility of being able to capture this data is especially probable since Denver International offers free WiFi and it is an unencrypted network.
It doesn't have to be unencrypted to be free.
Well, if you want a secure encrypted network, it's probably not going to be free.
There's only moderate additional security gained by having a WPA encrypted network where everyone has the same PSK since it's trivial to capture the association handshake (by forcing them to reassociate if neccessary) and steal the session key from anyone's session - Wireshark will do this for you. Alternatively, you can set up a hotspot on your laptop called "SouthwestAirlines" and nearby clients will connect to your laptop instead of the real Southwest network and you can capture all of their packets.
To make a secure encrypted network, they'd need to implement something like 802.1x security with unique username/passwords for each user and with Wifi clients configured to authenticate the network's 802.1x certificate (to prevent someone from setting up a rogue SouthwestAirlines access point).
Few providers of free Wifi service are going to be willing to run a helpdesk to assist all of the users with setting this up - it's not always trivial (depending on the device). So it's probably better to not provide the illusion of a secure encrypted network when it's not. The users that are sophisticated enough to set up 802.1x authentication on their device are probably also sophisticated to use a VPN to secure their data.
When I connect via an open Wifi network, I always VPN to my company or my home internet router so all of my wifi traffic is encrypted.