Neat idea, but I'd be concerned about cooling. The devices are all mounted upside-down, so the heat goes the "wrong" direction, and then they've got a closed surface above them for the warm air to pool and not vent.
Furthermore, it looks like more of these are contact burns from an iPod that's overheated, than are actual sparks flying - letalone spontaneous combustion. I've had laptops that (in normal operation) get nearly hot enough to burn skin.
You quibble over details when there's money to be made on lawsuits. Where are your priorities?
What's worrisome is that these same scientists who can't seem to build this thing without some fatal flaw are the same scientists telling us there's nothing to worry about when they create a black hole.
These are construction flaws. The fact that the black holes they may be able to create are not a threat has nothing to do with any sort of special containment.
Right, it just relies on them making no theoretical flaws.
The other "control test" which identifies everyone as a terrorist is [...] 0.00...001% accurate in identifying a terrorist.
Fail. Thanks for trying!
Millions of people who aren't terrorists, a few who are. A test that flags everyone as terrorists is therefore few/millions = 0.0000.....001% accurate. What is he missing?
Nope, there are other possibilities. For example, if 100% of the people screened aren't terrorists, then we don't know how well it works on terrorists, and do know that it identifies 10% of non-terrorists as terrorists. Buf if the people being screened are say 50% terrorists, then 90% correct result could mean it identifies all the non-terrorists correctly, and 80% of the terrorists, mis-identifying 20% of the terrorists (10% of the total people screened) as non-terrorists. etc. you get the idea. What we really want to know is what % of non-terrorists it flags as terrorists, and what % of terrorists if flags as non-terrorists (false-positive and false-negative, respectively).
So if it is 90% accurate in identifying terrorist vs. non-terrorist (and vice versa), then 10% of the non-terrorists will be identified as terrorists (or ~300),
You're assuming that the 10% erroneous operation identifies people as terrorists. If the 10% was always as non-terrorists, then it would flag nobody.
"To find one terrorist in 3000 people, using a screen that works 90% of the time, you'll end up detaining 300 people, one of whom might be your target."
If you have a screen that works 90% of the time, and you detain 300 people, 270 will be terrorists.
Let's say you've got no terrorists, the screening process works 90% of the time, and the other 10% it always flags people as terrorists. Then out of 3000, you'd get 300 flags that are all incorrect.
You should be alarmed! If you owned an iPod, you'd have a 1 in 11.7 million chance of having it catch fire spontaneously. That's more likely than dying due to being confined to or trapped in a low-oxygen environment!!!!
Lets be fair, he's in possession of stolen property
It's not stolen property because it wasn't stolen (or do you think all the people listed are now nameless and don't know why they are?). It's personal information, which can only be copied.
paint a STOLEN FROM... message on the front. Use indelible paint, or burn it into the surface of the netbook's plastic case.
Yes, this will be sure to make using the laptop yourself a pleasant experience anywhere public. Plus if you ever decide to sell it, it'll be sure to raise its resale value (especially if it's a Mac).
Re:The 2nd edition will use an unsigned byte count
on
The Geek Atlas
·
· Score: 1
The 2nd edition will use an unsigned byte count offering 256 cool places to visit.
Hand over your geek card; a signed (two's complement) octet can store up to 127, not 128.
BTW, it'd have been much more geeky if it said 0x80 places...
Myself, I prefer to draw the line at whatever the stock operating system provides. And then, they get to use everything they've got available to them. If "everything" happens to include the whole of DirectX, then so be it.
Sure, it's just that it's not as impressive when the OS provides really rich services. Having your code be able to access only the CPU, RAM, the video framebuffer, and a sample playback buffer, and still do interesting things is more impressive.
Meanwhile, every public pool has a policy of emptying everyone if thunder is heard. "Oh, you might get struck by lightening!" Yeah, well, you know what the chances of that are? A hell of a lot less than the risk that one of these brats is going to run out into the street and get run over by a car (perhaps while the driver is calling to see if the pool is open).
It's like people take all these precautions against the least likely dangers, while the more likely risks are ignored.
The pool situation is different because the person staying in the pool bears the risk, not others (unless they are required to pay for hospitalization, which wouldn't surprise me). With cellphone use, the driver is exposing everyone to any increased risk it causes. Big difference.
Passengers in the vehicle (at least those over 12) STFU where the driver is busy or when a situation develops, and their silence or their warnings actually calls attention to some dangers.
And you can understand a passenger without extensive mental effort. Sitting at home on a land line talking to my father on his cellphone, I have to devote almost all my attention to be able to understand what he says (and even then I often fail).
I've noticed that with my new touchscreen phone, even with text prediction it's a lot harder to do it because I have to look at the screen to make sure I'm hitting the right buttons, or at all. With my previous phone, I could just feel the buttons, and I knew what the text prediction would come up with, so I could write entire texts without looking at the phnoe until it was done.
And this is why "smart" (AKA unpredictable) computer interfaces are worse than dumb ones: you have to constantly find out what it did in response to your last input before you continue. It's essentially like a TCP connection with a very small window, preventing the sender from sending a lot at a time. Even though the "smart" interface might reduce the amount of input needed, it introduces lots of delays while you verify each step. Give me an interface I can simulate in my mind and therefore fire a lot of input at without constantly verifying that it's accepted as I expected.
The scary thing about this is that those numbers were from 2002. Think about how many more cellphones there are out there today than there were in 2002.
Hmmm... oh wait, need to change lanes... yeah people do use them... oh shit, gotta stop quickly, hold on... yeah, I bet it's highe
You said basically what I was going to say. Unlike the demos of a decade ago, these can use 3D libraries and music playback, so all this is doing is providing some 3D models (most likely algorithmically generated), textures, and some music sequence data (very likely using MIDI instrument sets that are built into the OS, or something similar). The ones from ages ago had their own 3D and music engines built in, outputting finished images to the framebuffer and sound to the playback buffer.
This also raises big problems as far as disease resistance goes -- if all the dogs are genetically identical they will all have identical immune systems, making it far easier for a single strain of disease to wipe out a large chunk of them.
And perhaps some weakness in their sniffing capabilities that could be exploited.
To me the problem is the Wikipedia rule of public use,' says Jerry Avenaim, a celebrity photographer. 'If they truly wanted to elevate the image on the site, they should allow photographers to maintain the copyright.'
Apparently they care more about freedom than having the highest quality images available. What more is there to say?
That's another problem with them; almost all are the damn 700MB instead of the proper 650MB. As I understand it, the 700MB discs pack the tracks closer together, so instead of them being spaced in the middle of the acceptable range, they are spaced right near the minimum accepted.
Slashdot Mirror
Neat idea, but I'd be concerned about cooling. The devices are all mounted upside-down, so the heat goes the "wrong" direction, and then they've got a closed surface above them for the warm air to pool and not vent.
One word: scissors. Took care of all my unsightly cabl
You quibble over details when there's money to be made on lawsuits. Where are your priorities?
I didn't realize that pigs had chimneys.
Right, it just relies on them making no theoretical flaws.
Millions of people who aren't terrorists, a few who are. A test that flags everyone as terrorists is therefore few/millions = 0.0000.....001% accurate. What is he missing?
Nope, there are other possibilities. For example, if 100% of the people screened aren't terrorists, then we don't know how well it works on terrorists, and do know that it identifies 10% of non-terrorists as terrorists. Buf if the people being screened are say 50% terrorists, then 90% correct result could mean it identifies all the non-terrorists correctly, and 80% of the terrorists, mis-identifying 20% of the terrorists (10% of the total people screened) as non-terrorists. etc. you get the idea. What we really want to know is what % of non-terrorists it flags as terrorists, and what % of terrorists if flags as non-terrorists (false-positive and false-negative, respectively).
You're assuming that the 10% erroneous operation identifies people as terrorists. If the 10% was always as non-terrorists, then it would flag nobody.
Let's say you've got no terrorists, the screening process works 90% of the time, and the other 10% it always flags people as terrorists. Then out of 3000, you'd get 300 flags that are all incorrect.
You should be alarmed! If you owned an iPod, you'd have a 1 in 11.7 million chance of having it catch fire spontaneously. That's more likely than dying due to being confined to or trapped in a low-oxygen environment!!!!
It's not stolen property because it wasn't stolen (or do you think all the people listed are now nameless and don't know why they are?). It's personal information, which can only be copied.
I found a fascinating audio explanation of how the guidance computer deals with deviation. Well worth the listen (only a couple of minutes long).
Yes, this will be sure to make using the laptop yourself a pleasant experience anywhere public. Plus if you ever decide to sell it, it'll be sure to raise its resale value (especially if it's a Mac).
Hand over your geek card; a signed (two's complement) octet can store up to 127, not 128.
BTW, it'd have been much more geeky if it said 0x80 places...
Sure, it's just that it's not as impressive when the OS provides really rich services. Having your code be able to access only the CPU, RAM, the video framebuffer, and a sample playback buffer, and still do interesting things is more impressive.
The pool situation is different because the person staying in the pool bears the risk, not others (unless they are required to pay for hospitalization, which wouldn't surprise me). With cellphone use, the driver is exposing everyone to any increased risk it causes. Big difference.
And you can understand a passenger without extensive mental effort. Sitting at home on a land line talking to my father on his cellphone, I have to devote almost all my attention to be able to understand what he says (and even then I often fail).
And this is why "smart" (AKA unpredictable) computer interfaces are worse than dumb ones: you have to constantly find out what it did in response to your last input before you continue. It's essentially like a TCP connection with a very small window, preventing the sender from sending a lot at a time. Even though the "smart" interface might reduce the amount of input needed, it introduces lots of delays while you verify each step. Give me an interface I can simulate in my mind and therefore fire a lot of input at without constantly verifying that it's accepted as I expected.
Hmmm... oh wait, need to change lanes... yeah people do use them... oh shit, gotta stop quickly, hold on... yeah, I bet it's highe
You said basically what I was going to say. Unlike the demos of a decade ago, these can use 3D libraries and music playback, so all this is doing is providing some 3D models (most likely algorithmically generated), textures, and some music sequence data (very likely using MIDI instrument sets that are built into the OS, or something similar). The ones from ages ago had their own 3D and music engines built in, outputting finished images to the framebuffer and sound to the playback buffer.
And perhaps some weakness in their sniffing capabilities that could be exploited.
DRM is dead, huh? Apparently Amazon didn't get the memo.
Apparently they care more about freedom than having the highest quality images available. What more is there to say?
That's another problem with them; almost all are the damn 700MB instead of the proper 650MB. As I understand it, the 700MB discs pack the tracks closer together, so instead of them being spaced in the middle of the acceptable range, they are spaced right near the minimum accepted.
Too late, the pirates already offer that.