In golf, you can display your wealth and not be any good at golf, and thats fine. They can see your a man of power becasue of the car you arrived in, the dimond tiped golf shoes and the caddie with a mobile drinks bar.
But in WoW if your new to the sport, or suck at it, you'll probably find that others who have less wealth and power in the real world can and probably will, have more in the World of Warcraft... egos don't like that
I thought that's what guild farmers were for. To let you turn real world wealth into in game prestige.:)
You too can buy the equivalent of those diamond tipped golf shoes.
For the sake of this response I'm going to assume that the parent post was serious, which may not have been the case.
While that design sounds like a nice simple idea, it does seem to overlook certain issues.
The first and possibly largest, is that it seems to assume that there is only one issue to be voted on. But on election days, there are often Senate and House races, local races, various propositions to be votes on, etc., in addition to the presidential election. (Assuming a presidential election year). The single dial can only handle a vote for a single race or issue. Adding a dial per race or issue gets ugly really quickly.
Two, requiring a voting token in the form of a post card invites people interested in performing vote tampering to either attempt to duplicate the vote token to allow multiple votes, or remove the vote token from the mail boxes of those people who they feel will vote against the issue. Or more simply, mis-delivered, late delivered tokens, or tokens that have been misplace since receiving them will also impact peoples ability to vote. (Of course you could add a back-up mechanism to handle missing tokens, but then that sacrifices much of the benefit of using tokens in the first place.)
Three, a rotary knob has an inherent maximum number of possible settings, and eventually some race will come along that has more candidates that whatever was chosen. (Or a vastly huge max selection will be chosen say 360, which will require excessive accuracy in selection, which could be an issue for older people).
Four, this system fails to account for those people with disabilities, either unable to read the printed names, or without the muscle control or gripping strength to manipulate the rotary wheel.
Five, this system does not appear to have a mechanism to allow a vote to be canceled and recast. So a mistaken vote would be permanent. (Adding a cancellation mechanism would greatly complicate the design of the unit).
Six, this system appears to be potentially vulnerable to malicious manipulation of the displayed selection for the rotary wheel. A false nameplate or sticker could be devised to cover the proper one, and scramble the positions of some of the candidates. (Possibly in an attempt to steal votes from a more popular candidate).
I'm sure that with a bit more though additional issues with this system could be realized. Voting system design is really easy to get wrong, because there are additional requirements that are often overlooked.
Alright everyone, I'm sure we can figure out how to solve this problem in our spare time between meetings and system rebuilds. After all, there's no problem NASA thinks is insurmountable that we can't convince ourselves is easily solved.
Sure its simple. At least for the stuff in fairly low orbit, which it was you care about.
Detonate large nukes just inside the atmosphere. This should create a bulge of atmosphere further into space than normal. As the objects run into this the higher atmospheric drag with cause them to deorbit more quickly.
The EMP burst and fallout are just minor implementation details. I'm sure engineering can have those solved next week.
So now for the important question, bagels or donuts for this meeting?
With one old version of Outlook, I started getting strange, scary, cryptic error messages. Then I noticed the folder they were going in had 16383 messages in it and figured they'd only allocated 14 bits somewhere.
I ran into that too. I failed to empty my deleted items for a couple months, and my spam rule failed because it could no longer delete the mail because deleted items was full.
(Not that the error message told you that. It was, as you said, cryptic)
Idiot: I started with a vintage Geiger Counter that I have had since high-school. It actually still worked too. Idiot: so I turned it into a sparkly piece of gayware why not!
Well aside from the fact that by the time one of those Lionel (yes the model train maker) Civil Defense Geiger counters reads anything you've already taken a really unhealthy, potentially lethal, dose of radiation you'd have had a point.
However, a "working" but unusable Geiger counter (heck I've got a couple of them around somewhere) or a slick looking iPod case.
So... we nearly say "Germany" instead of "Federal Republic of Germany", its real name. We use "Poland" instead of "Republic of Poland". We say "China" instead of "People's Republic of China". So, why won't we just call Taiwan... "Taiwan"?
Oh, I hated dealing with IRQs. Actually, I didn't mind so much until I had to deal with "Plug & Play" ISA hardware. I refused to buy anything that didn't let me use jumpers to force IRQs.
Sure most of the time P&P would actually manage to allocate everything an IRQ. But jeez was it dumb. It doesn't do me any good if my modem is on IRQ 11 and my sound card is on IRQ 15 (IDE chain 2 disabled). Because my terminal program expects the modem IRQ to match the standard one for its COM port, and game software wants the soundblaster to be on 7 or 11.
So sure, the system will boot, and windows will play sounds. But I don't have sound in my games or a usable modem. Great
Give me jumpers any day. I can force everything where I want it and where it will work.
But then PCI came along and largely fixed that. I say largely, because with enough slots you could get two card sharing a IRQ line when they couldn't share IRQs nicely. Sure you could assign that IRQ line any IRQ you wanted, but both cards got it. So you go off to play the PCI slot shuffle.
Yes. The Nimitz class have 2 reactors, and the Enterprise has 8
There are 9 Nimitz class carriers (18 reactors) and the Enterprise (8 reactors) for a total of 26. If you want to include non-US nuclear carriers (there is only one) the Charles De Gaulle has 2 reactors.
So worldwide there are 28 aircraft carrier nuclear reactors. Not 50 as claimed by a grandparent post.
Now I admit there are plenty more reactors on submarines (although, at least in all the US subs, only 1 reactor per sub). US subs only: Virigina Class SSN (1 in service) Seawolf Class SSN (3 in service) Los Angeles Class SSN (51 in service) Ohio Class SSBN (14 in service) Ohio Class SSGN (4 in service) So that is another 73 reactors.
Then there are other nations nuclear subs, there was one reactor powered cargo vessel, the US used to have some nuclear powered cruisers, the russian had several reactor powered nuclear icebreakers, etc.
So the overall point that there are a lot of Naval nuclear reactors is certainly correct.
The police officer is your accuser, and circumstantial evidence is not necessarily hearsay. Would you call DNA evidence hearsay? When someone robs a bank and gets caught on the security camera would you suggest we throw that evidence out as hearsay?
As much as I tend to agree with this I think there is a significant difference between an automated speeding camera and a bank security camera. (Or DNA)
In the case of the bank security camera it just records everything. Your accuser is the police officer or lab tech who reviewed the tapes, and determined that A) a crime had been committed, and B) that you were to perpetrator. DNA is a bit blurrier, since the computer might go ahead and generate the comparison between the two sets of DNA, but it would be the DNA lab tech's input that should determine if the match is close enough, so the lab tech would be the accuser.
For an automated speed camera, it isn't recording everyone, it is making a judgment based on sensors that someone is speeding. Being captured on camera is considered proof of speeding. Additionally, the picture usually doesn't have enough extra detail to independently verify if you were speeding or not. A single image can't be judged for speed. If it was a video tape then you could attempt to reconstruct the speed of the car and verify the camera's decision.
So in the case of an automatic speeding camera, the camera is the one deciding to accuse you (by taking the picture) and the police officer pretty much has to take the "word" of the camera.
I understand on some levels it's important to get the shuttle craft out of orbit, but since there is apparently a Soyuz capsule strapped to the ISS anyway, it might be that a safer solution would be to ride down to a hard landing on the proven Russian re-entry vehicle, which can later be returned to the ISS by rocket, and bring Discovery down on computer control.
Except that the Soyuz capsule can carry three, and then only if they have personally fitted acceleration seats. (Technically the seats are fixed, and each space station crewman carries up a form fitted insert for the seat to cushion them from the reentry forces).
There are currently 9 people on ISS, and except for the 2 assigned crewmen I don't believe any of them have the necessary seats for the Soyuz. So even if you wanted to use it only 1/3rd of the people could fit.
Fortunately protruding gap filler is a minor issue, because you just can't evacuate 9 people from ISS with the attached Soyuz.
Also consider the DC X ( or whatever it was called, Delta Clipper? ) which had such problems with cracks in the carbon fuel tanks and such. This stuff is *complicated*.
That was the Lockheed X-33 Venture Star. Where they were trying to build non spherical composite fuel tanks, and couldn't get the process right.
The DC-X was a subscale, sub orbital technical demonstration vehicle, which flew several times to show that vertical take-off vertical landing rocket powered vehicles were possible. At some point after the Air Force handed the project over to NASA the test vehicle was lost in a landing accident (a hydraulic line wasn't attached to one of the landing legs, so it landed and then fell over)
Many people believe that the DC-X based proposal would have been a better approach than awarding Lockheed the X-33 program, since the DC-X proposal was to build and fly subscale prototypes prior to attempting to build the single stage to orbit vehicle. That would have allowed problem to be discovered in testing and refinements made to the design.
Basically it wasn't as technically risky, as it required no significant new technology breakthroughs, while the X-33 required at least 2 (the linear aerospike engines, and the non-spherical composite fuel tanks). And even if the Delta Clipper (the name for the final SSTO vehicle) didn't work NASA still would have had significant flight test data from the DC-X and follow-on DC-Y vehicles. The DC-Y was expected to have a performance similar to what Rutan is calling for with Space Ship 2, exo-atmospheric flight with significant (minimum of several hundred mile) cross range capability.
All that said, I wouldn't dream of suggesting that Tivo Central Online and Tivo2Go come close to approximating the power of TivoWeb or the like. Tivo Online, for instance, has two fatal flaws:
It doesn't show you your current recording schedule
Recording are scheduled the next time your Tivo updates
As far as I'm concerned, those two flaws make the entire feature useless. The only time I would want remote scheduling is when I'm away from the Tivo (meaning I need some way of knowing what is scheduled) and forgot to record something (meaning I want it to take effect right NOW).
To be fair, assuming you have your TiVos connected to the internet (which I assume you do since you have them networked), the TiVo checks for new TiVo Online recording requests every 15 minutes. Not just every 24-36 hours when the daily call happens.
So you would have to be cutting it really close before using TiVo Online to schedule a recording was too slow.
(TiVos which are still stuck using their modem to dial in only check for recording requests during their daily calls)
Imagine this - you have a contract with your boss that any new inventions you make, you get 2.5% of the gross profit (remember no gross profit = no paycheck). You make your boss an invention that gets him 10 billion dollars. You are entitled to 250 million. They want to give you half of that because they think half is more then enough... How would you feel? Also remember, it was your reputation on the line.
Except this isn't what happened. A better version of this analogy would be:
Imagine this - you have a contract with your boss that any new inventions you make, you get 2.5% of the gross profit (remember no gross profit = no paycheck). You make your boss an invention that earns 12 billion dollars. Your boss contracts out the marketing and sale of this invention. A competative contract would give your boss 10 billion dollars, of which he would own you 250 million. But instead of getting the most competative contract, he awarded the contract to his wife and is only paid 5 billion, so he only pays you 125 million.
You are short 125 million, but he got the other 5 billion through his wife.
What you have apparently neglected to consider is that the reason that "the bends" are an issue is that it is difficult to carry enough O2 to decompress on the way up.
If you had essentially unlimited O2, then you could stay deeper for longer, and do proper decompression on the way up.
As for the pressure, the air is dissolved in the water, and hence is *already* at the same pressure as the water itself. No additional pressurization necessary.
Except that recreational SCUBA diving, like the grandparent post is referring to, is designed to avoid a decompression stage; both because it is an easy thing for recreational divers to forget to do / skimp on, and because it affects the ability to deal with any emergencies that might arise while underwater.
It's safer if you maintain a dive profile that always allows you to return straight to the surface.
So the fact that this device could allow you to maintain at 30 or 60 feet for the 30+ minutes it might take to safely decompress on the way up isn't likely to change the rules for recreational diving.
Now it may be a big advantage for commercial or military diving where the divers are professionals and are willing and able to do dives that require mandatory decompression stops..
(No matter how anonymous the card is, when you return to check out more books later, they can be waiting.)
Well, they could try to get you returning the books, since that would be a positive link between you and whatever "suspicious" book triggered this response. But they'd have to camp out on every branch library 24 hours a day (since I can return a book to any branch either at the counter or using the night book drop).
Otherwise the most information they could get from pulling the library's records would be that AnonymousCard serial number xyz, security deposit $50, was used to check out something they feel is suspicious.
Ok, with a live connection to the library's computers they could tell in real time when that card was used to check out new books. But they'd have to have someone on the scene to try to pick me up, so again we're back to stationing an officer in at least the likely branch libraries pretty much full time while they are open.
Of course, they could try to correlate the checkout against any security camera tapes, to attempt to get a picture and then use that to try to track you down.
That's getting back into a lot of manpower, which would tend to deter casual searches.
And if you are really paranoid you could always abandon a card, either temporarily or permanently, and pick up a new anonymous card. Imagine trying to track someone who checked out books once a month, but had 12 different anonymous cards. Rotated in order it would be a year before they reused a card.
One side effect of the scheme they are proposing is much lower portability of authentication data.
A username / password to connect to a website can be used from any browser that can connect to the website.
But using a salted hash like the SRP scheme they are talking about would require you either creating a new account from each browser you wanted to use, or moving the existing salt to each browser. Otherwise it wouldn't generate a matching hash, and would fail to verify. And creating multiple sets of authentication for one account probably wouldn't be popular with, for example, banks.
Now you could view this as a good thing, since it would discourage people from logging into secure web pages from untrustworthy computers (public terminals, etc.).
But it would also be a pain if you ever reinstalled your computer (oops, there went the salt values). Or wanted to log in to a page from more than one computer in your home, or from home and from work.
Admittedly the SRP scheme is only a small bit of this paper, and could be replaced, but it does seem like an unaddressed potential issue.
If I wanted blank keys I'd just pull the key caps off my IBM model M.
Price; $5 bucks at computer show. It's a decade old and going strong. I've got another in use elsewhere that is over 16 years old, still works perfectly.
Of course the other irritation with the film was the godawful dialogue. The "no I love you" "no I love you" scenes between Anakin and Padme, Vader lifting his head to the skies and shouting "NOOOOOOO!"
When I saw that I couldn't help thinking of Calculon. ("Funny story, the line was originally 'yes' in the script, but I put my own spin on it")
must admit, I did not realize that "Equilibrium" only came out on DVD, I always thought I missed it on the big screen, but now I think it is a gem.
It did release in theaters. Now apparently it wasn't a wide release, and I don't think it was advertised much, if at all. But I heard about it by word of mouth and saw it in the theater. Although if I recall correctly only one theater in the semi-local area was showing it.
Damn, that's subtle. I mean, I can figure out how it would work, but still. I didn't know it was possible to time anything as (ostensibly) quick as strcmp.
So, what's the solution? Adding a random delay to the loop?
Actually, just turning on md5 passwords (or SHA-1 passwords) would protect against this attack.
That is because those compare the hash of the entered password against a stored hash. Because those hashes are cryptographically strong one way functions knowing which is the first character not to match the stored hash doesn't provide you any help in guessing the correct password.
This strcmp timing attack only works against plaintext passwords. Well technically it should also work against passwords that were encrypted character by character, and work to a minor degree for passwords that were encrypted in multiple blocks. (You could tell which block was failing)
It seems to me that we should build the plants next to Yucca Mountain type facilities and use the national grid to transmit power to everywhere
The problem with that is nuclear plants really need access to a fair amount of water for cooling purposes; but long term storage facilities are focused on avoiding water, since it attacks the stored waste.
The geographic requirements for nuclear power plants and long term nuclear waste storage are just about opposite.
Why? The programmers lost a fight to fully automate the landing; but the code is in the machine. Just have the damn computer land the thing. It already applies the brakes! If I recall correctly, pretty much the only thing the pilot gets to do on landing a the shuttle is tell the computer to put the gear down. Maybe parent can confirm/deny this for me?:~)
Not sure about flight paths crossing over cities; I suppose that is probably the driving concern about tossing the shuttly in the water. That, and how would it look if the damn thing actually landed fine?;~)
Well, when Columbia broke up its debris certainly landed in a number of inhabited areas. It was pretty fortunate that no one on the ground was seriously injured.
However, I think that if the autopilot aimed for a landing at Edwards Air Force Base instead of Florida, then that risk could be minimized. That would put much more of the decent out over the Pacific ocean, so if that shuttle did break up it would have a much lower chance of injuring someone that it would if it broke up over Texas.
In addition the much longer and wider runway at Edwards would lower the risk of an automated landing. Sure it costs more to transport a shuttle from Edwards to Florida than to roll it off the runway and into the hanger, but its a heck of a lot cheaper than building a new shuttle.
The only sort of failure that would have them going to the ISS is something that would make impossible to land, such as damaged tiles. Any sort of life support system failure, they can still probably land the thing faster than they can dock to the station.
Of course there is a change of a failure that prevented both reentry and docking with ISS.
One off hand example might be explosive failure of one of the main engines. If it happened late enough in the flight the shuttle might well end up in an eccentric orbit, but with both tile damage and damage to the orbital maneuvering engines. Then it wouldn't be able to climb to ISS or reenter.
The only way a to save the astronauts would be to have a vehicle available that could match the damaged shuttles orbit and either space walk the astronauts across on a tether, or go for a hard docking and move the astronauts through a pressurized passage from the damaged shuttle to the rescue vehicle.
Unfortunately NASA doesn't have such a rescue vehicle.
One item I didn't see mentioned in the article is how the growing popularity of DVRs (be it TiVo, ReplayTV, MythTV, or Windows Media Center) might impact the view information gained from cable boxes.
Admittedly that is bypassed by the PPM that the article spent most of its time on, but viewer measurements from their cable boxes is something the article did mention.
Standalone DVRs (i.e. ones that aren't built into the cable or satellite box) appear to leave the cable box on continually. Even when they are recording it might be something equivalent to a TiVo suggestion, where a program is recorded in case the viewer might want to watch it. This would appear to really muck up any statistics pulled for such a cable box, since it is going to claim the viewer was watching something 24 hours a day, usually just whatever happens to be on the channel the last program was recorded from.
A few of these wouldn't throw the numbers off to badly, but the article was talking about 17 million DVRs by the end of the year. (Admittedly it wasn't broken down by integrated vs. standalone) That might be enough to mess up any scheme to measure audience through the cable box...
Slashdot Mirror
You too can buy the equivalent of those diamond tipped golf shoes.
For the sake of this response I'm going to assume that the parent post was serious, which may not have been the case.
While that design sounds like a nice simple idea, it does seem to overlook certain issues.
The first and possibly largest, is that it seems to assume that there is only one issue to be voted on. But on election days, there are often Senate and House races, local races, various propositions to be votes on, etc., in addition to the presidential election. (Assuming a presidential election year). The single dial can only handle a vote for a single race or issue. Adding a dial per race or issue gets ugly really quickly.
Two, requiring a voting token in the form of a post card invites people interested in performing vote tampering to either attempt to duplicate the vote token to allow multiple votes, or remove the vote token from the mail boxes of those people who they feel will vote against the issue. Or more simply, mis-delivered, late delivered tokens, or tokens that have been misplace since receiving them will also impact peoples ability to vote.
(Of course you could add a back-up mechanism to handle missing tokens, but then that sacrifices much of the benefit of using tokens in the first place.)
Three, a rotary knob has an inherent maximum number of possible settings, and eventually some race will come along that has more candidates that whatever was chosen. (Or a vastly huge max selection will be chosen say 360, which will require excessive accuracy in selection, which could be an issue for older people).
Four, this system fails to account for those people with disabilities, either unable to read the printed names, or without the muscle control or gripping strength to manipulate the rotary wheel.
Five, this system does not appear to have a mechanism to allow a vote to be canceled and recast. So a mistaken vote would be permanent. (Adding a cancellation mechanism would greatly complicate the design of the unit).
Six, this system appears to be potentially vulnerable to malicious manipulation of the displayed selection for the rotary wheel. A false nameplate or sticker could be devised to cover the proper one, and scramble the positions of some of the candidates. (Possibly in an attempt to steal votes from a more popular candidate).
I'm sure that with a bit more though additional issues with this system could be realized. Voting system design is really easy to get wrong, because there are additional requirements that are often overlooked.
Detonate large nukes just inside the atmosphere. This should create a bulge of atmosphere further into space than normal. As the objects run into this the higher atmospheric drag with cause them to deorbit more quickly.
The EMP burst and fallout are just minor implementation details. I'm sure engineering can have those solved next week.
So now for the important question, bagels or donuts for this meeting?
(Not that the error message told you that. It was, as you said, cryptic)
However, a "working" but unusable Geiger counter (heck I've got a couple of them around somewhere) or a slick looking iPod case.
I know which one gets you a Slashdot article.
Oh, I hated dealing with IRQs. Actually, I didn't mind so much until I had to deal with "Plug & Play" ISA hardware. I refused to buy anything that didn't let me use jumpers to force IRQs.
Sure most of the time P&P would actually manage to allocate everything an IRQ. But jeez was it dumb. It doesn't do me any good if my modem is on IRQ 11 and my sound card is on IRQ 15 (IDE chain 2 disabled). Because my terminal program expects the modem IRQ to match the standard one for its COM port, and game software wants the soundblaster to be on 7 or 11.
So sure, the system will boot, and windows will play sounds. But I don't have sound in my games or a usable modem. Great
Give me jumpers any day. I can force everything where I want it and where it will work.
But then PCI came along and largely fixed that.
I say largely, because with enough slots you could get two card sharing a IRQ line when they couldn't share IRQs nicely. Sure you could assign that IRQ line any IRQ you wanted, but both cards got it. So you go off to play the PCI slot shuffle.
There are 9 Nimitz class carriers (18 reactors) and the Enterprise (8 reactors) for a total of 26.
If you want to include non-US nuclear carriers (there is only one) the Charles De Gaulle has 2 reactors.
So worldwide there are 28 aircraft carrier nuclear reactors. Not 50 as claimed by a grandparent post.
Now I admit there are plenty more reactors on submarines (although, at least in all the US subs, only 1 reactor per sub).
US subs only:
Virigina Class SSN (1 in service)
Seawolf Class SSN (3 in service)
Los Angeles Class SSN (51 in service)
Ohio Class SSBN (14 in service)
Ohio Class SSGN (4 in service)
So that is another 73 reactors.
Then there are other nations nuclear subs, there was one reactor powered cargo vessel, the US used to have some nuclear powered cruisers, the russian had several reactor powered nuclear icebreakers, etc.
So the overall point that there are a lot of Naval nuclear reactors is certainly correct.
In the case of the bank security camera it just records everything. Your accuser is the police officer or lab tech who reviewed the tapes, and determined that A) a crime had been committed, and B) that you were to perpetrator.
DNA is a bit blurrier, since the computer might go ahead and generate the comparison between the two sets of DNA, but it would be the DNA lab tech's input that should determine if the match is close enough, so the lab tech would be the accuser.
For an automated speed camera, it isn't recording everyone, it is making a judgment based on sensors that someone is speeding. Being captured on camera is considered proof of speeding. Additionally, the picture usually doesn't have enough extra detail to independently verify if you were speeding or not. A single image can't be judged for speed. If it was a video tape then you could attempt to reconstruct the speed of the car and verify the camera's decision.
So in the case of an automatic speeding camera, the camera is the one deciding to accuse you (by taking the picture) and the police officer pretty much has to take the "word" of the camera.
There are currently 9 people on ISS, and except for the 2 assigned crewmen I don't believe any of them have the necessary seats for the Soyuz. So even if you wanted to use it only 1/3rd of the people could fit.
Fortunately protruding gap filler is a minor issue, because you just can't evacuate 9 people from ISS with the attached Soyuz.
The DC-X was a subscale, sub orbital technical demonstration vehicle, which flew several times to show that vertical take-off vertical landing rocket powered vehicles were possible. At some point after the Air Force handed the project over to NASA the test vehicle was lost in a landing accident (a hydraulic line wasn't attached to one of the landing legs, so it landed and then fell over)
Many people believe that the DC-X based proposal would have been a better approach than awarding Lockheed the X-33 program, since the DC-X proposal was to build and fly subscale prototypes prior to attempting to build the single stage to orbit vehicle. That would have allowed problem to be discovered in testing and refinements made to the design.
Basically it wasn't as technically risky, as it required no significant new technology breakthroughs, while the X-33 required at least 2 (the linear aerospike engines, and the non-spherical composite fuel tanks). And even if the Delta Clipper (the name for the final SSTO vehicle) didn't work NASA still would have had significant flight test data from the DC-X and follow-on DC-Y vehicles.
The DC-Y was expected to have a performance similar to what Rutan is calling for with Space Ship 2, exo-atmospheric flight with significant (minimum of several hundred mile) cross range capability.
So you would have to be cutting it really close before using TiVo Online to schedule a recording was too slow.
(TiVos which are still stuck using their modem to dial in only check for recording requests during their daily calls)
A better version of this analogy would be:
Imagine this - you have a contract with your boss that any new inventions you make, you get 2.5% of the gross profit (remember no gross profit = no paycheck). You make your boss an invention that earns 12 billion dollars.
Your boss contracts out the marketing and sale of this invention. A competative contract would give your boss 10 billion dollars, of which he would own you 250 million. But instead of getting the most competative contract, he awarded the contract to his wife and is only paid 5 billion, so he only pays you 125 million.
You are short 125 million, but he got the other 5 billion through his wife.
It's safer if you maintain a dive profile that always allows you to return straight to the surface.
So the fact that this device could allow you to maintain at 30 or 60 feet for the 30+ minutes it might take to safely decompress on the way up isn't likely to change the rules for recreational diving.
Now it may be a big advantage for commercial or military diving where the divers are professionals and are willing and able to do dives that require mandatory decompression stops..
But they'd have to camp out on every branch library 24 hours a day (since I can return a book to any branch either at the counter or using the night book drop).
Otherwise the most information they could get from pulling the library's records would be that AnonymousCard serial number xyz, security deposit $50, was used to check out something they feel is suspicious.
Ok, with a live connection to the library's computers they could tell in real time when that card was used to check out new books. But they'd have to have someone on the scene to try to pick me up, so again we're back to stationing an officer in at least the likely branch libraries pretty much full time while they are open.
Of course, they could try to correlate the checkout against any security camera tapes, to attempt to get a picture and then use that to try to track you down.
That's getting back into a lot of manpower, which would tend to deter casual searches.
And if you are really paranoid you could always abandon a card, either temporarily or permanently, and pick up a new anonymous card. Imagine trying to track someone who checked out books once a month, but had 12 different anonymous cards. Rotated in order it would be a year before they reused a card.
One side effect of the scheme they are proposing is much lower portability of authentication data.
A username / password to connect to a website can be used from any browser that can connect to the website.
But using a salted hash like the SRP scheme they are talking about would require you either creating a new account from each browser you wanted to use, or moving the existing salt to each browser. Otherwise it wouldn't generate a matching hash, and would fail to verify.
And creating multiple sets of authentication for one account probably wouldn't be popular with, for example, banks.
Now you could view this as a good thing, since it would discourage people from logging into secure web pages from untrustworthy computers (public terminals, etc.).
But it would also be a pain if you ever reinstalled your computer (oops, there went the salt values). Or wanted to log in to a page from more than one computer in your home, or from home and from work.
Admittedly the SRP scheme is only a small bit of this paper, and could be replaced, but it does seem like an unaddressed potential issue.
If I wanted blank keys I'd just pull the key caps off my IBM model M.
Price; $5 bucks at computer show. It's a decade old and going strong. I've got another in use elsewhere that is over 16 years old, still works perfectly.
But I heard about it by word of mouth and saw it in the theater. Although if I recall correctly only one theater in the semi-local area was showing it.
That is because those compare the hash of the entered password against a stored hash. Because those hashes are cryptographically strong one way functions knowing which is the first character not to match the stored hash doesn't provide you any help in guessing the correct password.
This strcmp timing attack only works against plaintext passwords.
Well technically it should also work against passwords that were encrypted character by character, and work to a minor degree for passwords that were encrypted in multiple blocks. (You could tell which block was failing)
The geographic requirements for nuclear power plants and long term nuclear waste storage are just about opposite.
However, I think that if the autopilot aimed for a landing at Edwards Air Force Base instead of Florida, then that risk could be minimized. That would put much more of the decent out over the Pacific ocean, so if that shuttle did break up it would have a much lower chance of injuring someone that it would if it broke up over Texas.
In addition the much longer and wider runway at Edwards would lower the risk of an automated landing. Sure it costs more to transport a shuttle from Edwards to Florida than to roll it off the runway and into the hanger, but its a heck of a lot cheaper than building a new shuttle.
One off hand example might be explosive failure of one of the main engines. If it happened late enough in the flight the shuttle might well end up in an eccentric orbit, but with both tile damage and damage to the orbital maneuvering engines. Then it wouldn't be able to climb to ISS or reenter.
The only way a to save the astronauts would be to have a vehicle available that could match the damaged shuttles orbit and either space walk the astronauts across on a tether, or go for a hard docking and move the astronauts through a pressurized passage from the damaged shuttle to the rescue vehicle.
Unfortunately NASA doesn't have such a rescue vehicle.
One item I didn't see mentioned in the article is how the growing popularity of DVRs (be it TiVo, ReplayTV, MythTV, or Windows Media Center) might impact the view information gained from cable boxes.
Admittedly that is bypassed by the PPM that the article spent most of its time on, but viewer measurements from their cable boxes is something the article did mention.
Standalone DVRs (i.e. ones that aren't built into the cable or satellite box) appear to leave the cable box on continually. Even when they are recording it might be something equivalent to a TiVo suggestion, where a program is recorded in case the viewer might want to watch it.
This would appear to really muck up any statistics pulled for such a cable box, since it is going to claim the viewer was watching something 24 hours a day, usually just whatever happens to be on the channel the last program was recorded from.
A few of these wouldn't throw the numbers off to badly, but the article was talking about 17 million DVRs by the end of the year. (Admittedly it wasn't broken down by integrated vs. standalone) That might be enough to mess up any scheme to measure audience through the cable box...