There was a bug in one vendor's support of bluetooth that allowed it to accept things without proper authentication/confirmation. One still had to RUN the program MANUALLY before it would 'infect' you and begin attempting to infect others.
Yes, but it was coupled with a stupid design that virtually guaranteed that many would fall for it;
Alertbox comes up: "Do you want to accept ? [Yes] [No]"
User selects "No"
Two seconds later (since the infected device is still in range) the same alertbox comes again. And again. And again.
People are used to those, they'll quickly get the idea that "No" doesn't work and only clicking "Yes" will make the pesky alertbox go away (which is what they want, they don't care what the box says or why it's there, they only want it to go away so they can go on with whatever they where doing)
Yeah, it's a lot of money. At the moment around 8 trillion (8*10^12) dollars. And it's growing at an alarming rate (alarming as in much quicker than the BNP of the USA)
I agree with you that it very likely won't be paid back. But on the other hand I'm not sure it won't stabilise (as in start to grow no more than the BNP) which would make it indefinitely sustainable.
It's about 25K for every American, at todays interest-rates this means every American is out around 750$ a year in interest.
That's a lot of money, but it's still a small part of the average american income of $40.000. 2% of Americas income goes to pay interest on the loans.
If the wild growth doesn't slow or stop though, or if the interest-rates grow, it's soon going to be quite a lot more.
Unless you've got a Hauppage PVR350 or similar card that does hardware mpeg-compression. If you do you need basically no cpu at all. (I get my 900mhz low-power Via Neremiah to 5% load or so, but that's a very weak cpu)
The storage-requirements would still be humongous though, recording *one* channel 24/7 requires on the order of 1TB/month, so you'd like 4*250GB drives for each channel. (yes, there's bigger disks, but they cost *more* for the same storage)
Like I said: for that problem to arise you'd have to have really stupid friends.
There is indeed a lot of things you are not allowed to do with proprietary software. No, users aren't physically chained down, but friends with an iq above room-temperature would in general be able to understand freedom as a *sligthly* more generalised concept than freedom from chains.
If your friends honestly, after you tried to explain it are unable to understand the distinction between freedom and gratis, then I think you need new friends, or work on your explanation-skills.
I hear this all the times -- from people who *have* understood it -- that there's, supposedly, hordes of people who have problems grasping this distinction. Comically enough you never seem to meet people who don't understand it. Only people who claim that lots of *others* don't understand it.
You "realize" it takes "a lot more" electricity to run A/C than heat ?
Why, exactly, would that be ?
The most efficient way of heating with electricity is to use a heat-pump. A heat-pump and a A/C unit is *exactly* the same thing, only one is working in reverse relative to the other.
Cooling takes more than heating only if your house is so constructed that it's better at keeping heat in than at keeping it out. That may be the case, for example, if you've got large modern low-heatflow windows. Those let the sun in no problem, but let very little heat back out. (this is a good thing if you live somewhere where it's more often too cold than too hot)
Given reasonable construction for the climate you're in there's not a single reason why cooling should require more than heating, assuming the temperature-differential you want to maintain is the same.
In reality, mostly the temperature-differential is a lot smaller.
I realize you're (attempting to:-) be funny, but it really is not very complicated. Unless you've got really braindead friends it should be possible to get the idea across to them that the main idea is freedom, as in the oposite of slavery, the oposite of imprisonment, the oposite of being restrained. And not "free" as in "at no monetary cost".
But humanity is by far not separated enough for this to happen. It doesn't take that large a percentage of cross-overs for the gene-pools to stay united.
And you forget that desirability is influenced by a lot more than looks, especially for males.
Some ugly male end up the leading singer of a popular rock-band, and suddenly there's lots of "hotties" interested in him.
Or he simply ends up with a lot of money, for whatever reason. Or terribly good in sports, or any other of the hundreds of things someone finds attractive.
Why would anyone want to increase latency anyway ?
You're right btw, raid doesn't increase latency, actually depending on what kind of raid and what kind of access-patterns it can even decrease average latency.
You're rigth, in principle. But in practice I think the personal-image (or the more or less equivalent "security skin") would be helpful:
Setting up a full MitM website is quite a bit more complicated than just a simple static website with a form. I've seen some of the forms even use mailtos as action, freeing the phisher from needing any server-side scripting.
With one-time passwords (for example the tans common in Europe) MitM has only limited use, because the tan used for authorising one transaction is useless thereafter.
If the bank uses personal certificates, like Skandiabanken and others here have been doing for literally years, the trick won't work. Your browser won't present even the public part of your "skandiabanken.no" certificate to "phishersite.org", much less the secret part. Thus the phishersite *can't* successfully do a MitM.
There's no excuse not to use the two-way authenthication built into SSL really. All major browsers have supported it for years, it is very simple to use and imposes no additional hassle on the user, and it adds *significantly* to the security of an online bank.
I don't mean to use *only* this, the certificate could be stolen if, for example, someone installed a trojan on your machine.
But the current combo, as used commonly in Europe gives a quite decent overall protection:
Personal SSL-certificate -- identifies you to the bank. Prevents MitM. Prevents login by people who somehow knows your userid/pin. Allows personalised login-page which makes things harder for phishers.
Normal pin -- prevents logins from someone having say stolen your computer, or from someone you've let borrow your computer.
transaction-authorisation-number (tan) -- limits the usefulness of MitM attacks (even if the certificate didn't prevent them. Stops even a person having a rootkit on your computer. He can get your pin by keysniffing, and your certificate from the disk, but he won't have physical access to your tan-list or tan-generating-device.
This migth sound cumbersome to the customer, but infact it isn't. For checking saldos or whatever from home all you have to do is enter userid and pin. For doing transactions you need to enter a tan aditionally. The personal SSL-certificate you don't even notice after you install it on your first bank-visit.
The volume of air in the car is only marginally interesting when calculating the needed cooling.
To keep a car cool, the AC has to remove the same amount of heat as are being added by the sun/outside temperature, persons in the vehicle etc. Plus a bit so that it'll be capable of *cooling* the car and not only capable of *keeping* it cool.
The average house does not consist of single windows and uninsulated metal all around. A typical house is insulated, which works just as well in keeping heat out as in keeping it in. The average car is not. (or very little)
100BTU/hour is about 293 Watts (like so many nonmetric units, BTU is poorly defined, it's value varies somewhat depending on temperature), so a system with an COP of 0.4 would need to consume 293/0.4 = 732W to cool that much.
This is certainly possible, but it's a lot of power for a car, at the voltages used you'd need more than 60A.
A compressor-cooler with a COP of 3 on the other hand will require about 100W, which works out to around 8A.
Perhaps there's a reason why this is the solution used by all current auto-makers ?
This seems like a combination of the typical insecure, stupid "personal question" with an actually good idea: the personal image.
The first, using a "personal question" as a means of making easily guessable passwords more secure is dumb. It is true that people often choose easily guessable passwords. But people *even* more often choose easily guessable "personal questions". "Mothers maiden name" for example. That's how Paris Hiltons adress-book got cracked: She'd used the hugely difficult "personal question" about the name of her dog. It takes only 10 seconds of googling to find the answer to that...
The personally selected secret image on the other hand is a good idea: phishers rely on the fact that they can easily create a fake website that looks like the real one.
If the real one has some element that is unique to you, they won't be able to copy that, simply because they don't know what it is.
This *ain't* the system common in Scandinavia (and other countries) by the way. What we have is generally a one-time "tan" to authorise transactions, provided either as a paper-list where you cancel out those you used, or from a small cryptographic device that generates them using the current time, your account-number and a secret embedded key.
It is, however, just a weaker version of the proposed "security skins", which is an excellent idea to prevent or reduce phishing.
My bank, Skandiabanken does this, sort of, already. (though they underpublizise it). There each user has a private security-certificate used to authenticate the user, in addition to the pin.
This helps in two ways:
First, even if you knew my customer-id and my pin, you still could not log in on my account, you wouldn't have the certificate.
Secondly, it enables the bank to identify me even before I log in, thus giving me a personal greeting not easily copied by phishers: on the login page, before I've entered anything the bank says: "Hello Eivind Kjørstad."
Phishers have no easy way of doing that, they generally don't have a clue which user is sitting behind which ip.
What computers will do 20 years from now, obviously none of us can answer, if we could, we'd be out there making it happen.
I can make a comparison to the C64 though, in that case the computer also does fundamentally new things, not just faster-prettier versions of the same-old.
Notice that I say *do* not "can" do. A C64 was physically capable of doing many of those things, but that doesn't change the fact that the typical one didn't.
The biggie is obviously networking and the Internet. There are people today that have computers solely for the reason that they want to use the Internet. (My grandmother for a start) It was physically possible to network C64-machines, but I'd wager a bet that most people didn't, not even BBS-type networking.
I/O. It's practically possible, and indeed common to use a modern day computer to store all your pictures from your digital camera. The C64 could not and did not store thousands of pictures at acceptable quality. You could argue this is just more/faster, but infact it's a field of use that just wasn't there with the C64.
Multimedia. The C64 could not reasonable show video, or play music (sids are/music/ but they never did replace the LP, for very understandable reasons)
Services. I can, and do, use my computer for filing my tax return, for ordering pizza, for paying my bills, for selling my old c64-game-collection, for ordering plane-tickets, for sending pictures of her great-grandson to my grandmother. None of which was practically possible with the C64. (many of these where *theoretically* possible)
You can argue that all this is faster/more. But sometimes, when a quantitative difference becomes large enough it becomes a qualitative difference, because things become possible that just wheren't realistic before.
Now, what will come in the next 20 years is anyones guess.
speed, storage, size, cost, i/o and connectivity is more like it, and even then I'm sure I've forgotten something.
A Eniac with more speed and storage, but the same cost as it had back then would not have been a modern PC, you wouldn't have them in homes for a start.
Same goes if it was still the size it had back then. Or if it still used the same I/O, or if it lacked the connectivity of modern computers.
My first computer was a C64. It had 64 kilobytes of ram, a tape-deck and a 4mhz cpu.
Today, about 20 years later my computer has 8000 times that amount of memory and a clockspeed 1000 times higher, while operating on 4 times as large units.
It also cost significantly *less* than the C64 once did. (though I got mine used for cheap)
If this trend continues, by the time I'm 50 (another 20 years) A new (but not high-end) desktop-computer will have 4 terabytes of RAM, operate at 4 teraherz with 256bit operands and have more disk-space than Google today.
That's pretty hard to believe, for one at that clockspeed *light* itself manages only 0.25mm so it'd be physically impossible unless the chips get a lot smaller. (connections longer than that couldn't propagate the signal in time)
Still, even if we don't get that far it'll be interesting to see how far we *do* get. The only thing I'm sure of is that the computer I'll have when I turn 50 will make anything that exists today look like a toy.
How, exactly would this be different if the college kid or the woman made the publication on something old-fashioned like oh, say, paper ?
Can women who pose nude in magazines "undo" it after they got their money if it doesn't suit them anymore ?
Can Newspapers that wrote something positive about oh, say, Bin Laden 20 years ago "erase" it, and ensure that noone learns of it ?
Face it: The moment you do something, it's done. It can't be *undone*, nor *should* that be possible.
Instead people should consider actually either standing for what they've done, or alternatively not do it. There's no third alternative. Not on the web, not anywhere else.
The problem with this theory is that your body is evolved to have the largest possible chanse of having as many as possible healthy offspring under certain circumstances. This does not include you being healthy or living to a high age.
Evolution doesn't help you live healthily to a higher age than the one where you can benefit your children. It may even be *beneficial* for your genes that you die the moment you're more a competitor for your children than an assistance for them.
Evolution only works aslong as a certain trait helps you have children, or helps your children have children.
Science is about repeatable controlled experiments that yield consistent results. Repeatable means that you need to understand what exactly is going on in your experimental setup so that somebody else can reproduce it. Controlled means you account for all variables and only vary one at a time. The problem is that doing all this correctly with people costs a LOT of money.
Not only. The problem is that given that any two people are different in a million ways that it's simple not possible.
You can *never* have two groups of people that are identical, except for one factor (which is the one you're interested in.)
Yeah, sure, you can try various aproximations of this ideal, and given enough of a budget, you can sometimes get reasonably close. But you can never achieve it. You just have to do your best, and then hope that whatever other, unaccounted for, differences doesn't mess up your result.
That's true for all experiments in the real world really, but it's *more* true for people than for say spheres of lead falling in vacuum.
Particularily with problems that are uncommon or rare it's a huge problem to get enough of a sample-size that there's still any sample left after you correct for the obvious and/or known factors.
Take SIDS in Europe for example. Incidence is 1:5000 or thereabout. We know that smoking increases the risk quite a lot, so any study that wants to do research on *other* factors needs to factor for smoking or non-smoking.
We also know that low birth-weigtth, young mothers, insufficient pre-natal care and certain sleeping-conditions have an effect. Once you split for all of these, you'd need a gargantuan start-population to have anything left at all.
1:5000 children will die of SIDS. 1:3 children have parents that smoke. (2:3 of the SIDS-dead) 3% of all children have low birth-weigth. etc.
If you wanted to do research on other theories, say the theory that gases given of from certain foam matresses play a role, you'd need to have initial data for literally millions of children to have even a *hope* of correctly canceling all those other effects an narrow in on the influence (or noninfluence) of mattresses.
As if this wasn't bad enough:
PArticipation in medical experiments is generally voluntarily, and some people refuse or drop out during the experiment or whatever. But and that's the tricky part -- the ones dropping out are *NOT* a random part of the population, but rather a certain type of people, so this f*cks up your data too and needs to be corrected, as good as you manage.
Correct. And the protection you get is in a certain sense inversely proportional to the uniqueness of the trademark.
If I create the trademark "Quropiwla", which is a freely invented word, then I'm going to have a fair chanse demonstrating that anyone using that word does it in the intention of creating confusion with my product -- afterall, what other reason could there be ? It's unlikely someone would end up with that word by chance, and the word doesn't have any meaning as such.
If, on the other hand I register "Enjoyable" in the market for computer-input-devices (say I market joysticks) then it's not that obvious that others using this word does so in order to create confusion with my product. I'd probably still be able to stop other "Enjoyable" joysticks, but I doubt I'd have a case against say the "Enjoyable" computer-monitor.
If a rotor on a moving craft moves quickly enough that the tip-speed is equal to the speed of the craft, then the tip of the rotor will stand still, relative to the surrounding air on its way backwards.
But all other parts of the rotor move *slower* than the tip, so no part of the rotor will move backwards relative to the surrounding air. That's just bullshit.
Fine. But we're not talking existing servers here, but rather new ones (the cards are not even on the market yet).
So the question is, how many of the ones that buy a new server *after* these cards become available on the market will install a kernel older than 2.6.10?
Perhaps a few, but probably *not* the ones for which performance is important, nor the majority which simply takes whatever is default in their distro.
Fedora and Mandriva both have newer default kernels (2.6.11 I think), and I guess even Debian uses 2.6 now ?
Smart cards do more than just offload the CPU from handling TCP data. They offload the PCI bus from handling interrupts. Every packet of data triggers an interrupt.
With newer Linux-kernels this is quite simply not the case.
To avoid the torrent of interupts from a fast nic the Linux-kernel detects that the card gets packets so often that essentially there's "always" one or more packets waiting in the cards buffers.
It responds to this condition by disabling interupts for the card in question and switch to polling it regularily.
Normally polling is inefficient, because it amounts to asking over and over again "got anything for me now?", where in most situations the answer is "no" 99.99% of the time, which makes it a waste of resources to ask in the first place.
This changes when the answer is "yes" basically all the time. There's no need for the network card to tell the cpu over and over and over "I got a packet for you", instead the cpu collects packets regularily.
It's sorta like having your lawyer receive legal letters for you.
If you get very few, it'd be a waste for you to drop by him every day and ask if he's gotten any for you (polling) most of the days you'd be making the trip for naught. In this situation interupts (i.e. having your lawyer call you and inform you on the rare occasions when a letter *does* arrive) makes more sense.
But if your letter-flow increases to the point where there's normally 3-5 letters every day and it's rare that no letter arrives, then it no longer makes sense that the lawyer calls you every day to tell you you got letters (interupts), you already assumed that. In this scenario it makes more sense for you to come by regularily and pick up letters without being prompted to do so. (polling)
Thus, the flow of interupts from a Gb-nic being flooded with 100byte small-packets (say a loaded dns-server) is not 1 million interupts every second -- it is zero interupts.
(allthough what you write is correct for kernels older than 2.6.10 and for less clever OSes.)
You're -again- assuming that everyone is and uses englisch exclusively. That's simply not the case.
Sure, your "fix" sort of helps. A little. For those using english. Assuming they're using the "correct" font (some glyphs look similar or not depending on the fonts used)
"No good reason" to use international domain-names in a trustworthy domain name ? Isn't that the same as saying there's no good reason to use them at all ?
The *obvious* use of international domain names is for companies to be able to use the domain name corresponding to their actual name.
If you can't do that (reliably) there's no (or very little) point in having international domain names at all. Your "fix" thus isn't a "fix" unless you consider: remove all traces of idn to be a "fix".
Slashdot Mirror
Yes, but it was coupled with a stupid design that virtually guaranteed that many would fall for it;
People are used to those, they'll quickly get the idea that "No" doesn't work and only clicking "Yes" will make the pesky alertbox go away (which is what they want, they don't care what the box says or why it's there, they only want it to go away so they can go on with whatever they where doing)
I agree with you that it very likely won't be paid back. But on the other hand I'm not sure it won't stabilise (as in start to grow no more than the BNP) which would make it indefinitely sustainable.
It's about 25K for every American, at todays interest-rates this means every American is out around 750$ a year in interest.
That's a lot of money, but it's still a small part of the average american income of $40.000. 2% of Americas income goes to pay interest on the loans.
If the wild growth doesn't slow or stop though, or if the interest-rates grow, it's soon going to be quite a lot more.
Doing everything in your list to perfection still is pretty useless if you haven't got atleast basic physical security.
The storage-requirements would still be humongous though, recording *one* channel 24/7 requires on the order of 1TB/month, so you'd like 4*250GB drives for each channel. (yes, there's bigger disks, but they cost *more* for the same storage)
There is indeed a lot of things you are not allowed to do with proprietary software. No, users aren't physically chained down, but friends with an iq above room-temperature would in general be able to understand freedom as a *sligthly* more generalised concept than freedom from chains.
If your friends honestly, after you tried to explain it are unable to understand the distinction between freedom and gratis, then I think you need new friends, or work on your explanation-skills.
I hear this all the times -- from people who *have* understood it -- that there's, supposedly, hordes of people who have problems grasping this distinction. Comically enough you never seem to meet people who don't understand it. Only people who claim that lots of *others* don't understand it.
Why, exactly, would that be ?
The most efficient way of heating with electricity is to use a heat-pump. A heat-pump and a A/C unit is *exactly* the same thing, only one is working in reverse relative to the other.
Cooling takes more than heating only if your house is so constructed that it's better at keeping heat in than at keeping it out. That may be the case, for example, if you've got large modern low-heatflow windows. Those let the sun in no problem, but let very little heat back out. (this is a good thing if you live somewhere where it's more often too cold than too hot)
Given reasonable construction for the climate you're in there's not a single reason why cooling should require more than heating, assuming the temperature-differential you want to maintain is the same.
In reality, mostly the temperature-differential is a lot smaller.
And you forget that desirability is influenced by a lot more than looks, especially for males.
Some ugly male end up the leading singer of a popular rock-band, and suddenly there's lots of "hotties" interested in him.
Or he simply ends up with a lot of money, for whatever reason. Or terribly good in sports, or any other of the hundreds of things someone finds attractive.
You're right btw, raid doesn't increase latency, actually depending on what kind of raid and what kind of access-patterns it can even decrease average latency.
There's no excuse not to use the two-way authenthication built into SSL really. All major browsers have supported it for years, it is very simple to use and imposes no additional hassle on the user, and it adds *significantly* to the security of an online bank.
I don't mean to use *only* this, the certificate could be stolen if, for example, someone installed a trojan on your machine.
But the current combo, as used commonly in Europe gives a quite decent overall protection:
This migth sound cumbersome to the customer, but infact it isn't. For checking saldos or whatever from home all you have to do is enter userid and pin. For doing transactions you need to enter a tan aditionally. The personal SSL-certificate you don't even notice after you install it on your first bank-visit.
To keep a car cool, the AC has to remove the same amount of heat as are being added by the sun/outside temperature, persons in the vehicle etc. Plus a bit so that it'll be capable of *cooling* the car and not only capable of *keeping* it cool.
The average house does not consist of single windows and uninsulated metal all around. A typical house is insulated, which works just as well in keeping heat out as in keeping it in. The average car is not. (or very little)
100BTU/hour is about 293 Watts (like so many nonmetric units, BTU is poorly defined, it's value varies somewhat depending on temperature), so a system with an COP of 0.4 would need to consume 293/0.4 = 732W to cool that much.
This is certainly possible, but it's a lot of power for a car, at the voltages used you'd need more than 60A.
A compressor-cooler with a COP of 3 on the other hand will require about 100W, which works out to around 8A.
Perhaps there's a reason why this is the solution used by all current auto-makers ?
The first, using a "personal question" as a means of making easily guessable passwords more secure is dumb. It is true that people often choose easily guessable passwords. But people *even* more often choose easily guessable "personal questions". "Mothers maiden name" for example. That's how Paris Hiltons adress-book got cracked: She'd used the hugely difficult "personal question" about the name of her dog. It takes only 10 seconds of googling to find the answer to that...
The personally selected secret image on the other hand is a good idea: phishers rely on the fact that they can easily create a fake website that looks like the real one.
If the real one has some element that is unique to you, they won't be able to copy that, simply because they don't know what it is.
This *ain't* the system common in Scandinavia (and other countries) by the way. What we have is generally a one-time "tan" to authorise transactions, provided either as a paper-list where you cancel out those you used, or from a small cryptographic device that generates them using the current time, your account-number and a secret embedded key.
It is, however, just a weaker version of the proposed "security skins", which is an excellent idea to prevent or reduce phishing.
My bank, Skandiabanken does this, sort of, already. (though they underpublizise it). There each user has a private security-certificate used to authenticate the user, in addition to the pin.
This helps in two ways:
First, even if you knew my customer-id and my pin, you still could not log in on my account, you wouldn't have the certificate.
Secondly, it enables the bank to identify me even before I log in, thus giving me a personal greeting not easily copied by phishers: on the login page, before I've entered anything the bank says: "Hello Eivind Kjørstad."
Phishers have no easy way of doing that, they generally don't have a clue which user is sitting behind which ip.
I can make a comparison to the C64 though, in that case the computer also does fundamentally new things, not just faster-prettier versions of the same-old.
Notice that I say *do* not "can" do. A C64 was physically capable of doing many of those things, but that doesn't change the fact that the typical one didn't.
- The biggie is obviously networking and the Internet. There are people today that have computers solely for the reason that they want to use the Internet. (My grandmother for a start) It was physically possible to network C64-machines, but I'd wager a bet that most people didn't, not even BBS-type networking.
- I/O. It's practically possible, and indeed common to use a modern day computer to store all your pictures from your digital camera. The C64 could not and did not store thousands of pictures at acceptable quality. You could argue this is just more/faster, but infact it's a field of use that just wasn't there with the C64.
- Multimedia. The C64 could not reasonable show video, or play music (sids are
/music/ but they never did replace the LP, for very understandable reasons)
- Services. I can, and do, use my computer for filing my tax return, for ordering pizza, for paying my bills, for selling my old c64-game-collection, for ordering plane-tickets, for sending pictures of her great-grandson to my grandmother. None of which was practically possible with the C64. (many of these where *theoretically* possible)
You can argue that all this is faster/more. But sometimes, when a quantitative difference becomes large enough it becomes a qualitative difference, because things become possible that just wheren't realistic before.Now, what will come in the next 20 years is anyones guess.
A Eniac with more speed and storage, but the same cost as it had back then would not have been a modern PC, you wouldn't have them in homes for a start.
Same goes if it was still the size it had back then. Or if it still used the same I/O, or if it lacked the connectivity of modern computers.
My first computer was a C64. It had 64 kilobytes of ram, a tape-deck and a 4mhz cpu.
Today, about 20 years later my computer has 8000 times that amount of memory and a clockspeed 1000 times higher, while operating on 4 times as large units.
It also cost significantly *less* than the C64 once did. (though I got mine used for cheap)
If this trend continues, by the time I'm 50 (another 20 years) A new (but not high-end) desktop-computer will have 4 terabytes of RAM, operate at 4 teraherz with 256bit operands and have more disk-space than Google today.
That's pretty hard to believe, for one at that clockspeed *light* itself manages only 0.25mm so it'd be physically impossible unless the chips get a lot smaller. (connections longer than that couldn't propagate the signal in time)
Still, even if we don't get that far it'll be interesting to see how far we *do* get. The only thing I'm sure of is that the computer I'll have when I turn 50 will make anything that exists today look like a toy.
Can women who pose nude in magazines "undo" it after they got their money if it doesn't suit them anymore ?
Can Newspapers that wrote something positive about oh, say, Bin Laden 20 years ago "erase" it, and ensure that noone learns of it ?
Face it: The moment you do something, it's done. It can't be *undone*, nor *should* that be possible.
Instead people should consider actually either standing for what they've done, or alternatively not do it. There's no third alternative. Not on the web, not anywhere else.
Evolution doesn't help you live healthily to a higher age than the one where you can benefit your children. It may even be *beneficial* for your genes that you die the moment you're more a competitor for your children than an assistance for them.
Evolution only works aslong as a certain trait helps you have children, or helps your children have children.
Not only. The problem is that given that any two people are different in a million ways that it's simple not possible.
You can *never* have two groups of people that are identical, except for one factor (which is the one you're interested in.)
Yeah, sure, you can try various aproximations of this ideal, and given enough of a budget, you can sometimes get reasonably close. But you can never achieve it. You just have to do your best, and then hope that whatever other, unaccounted for, differences doesn't mess up your result.
That's true for all experiments in the real world really, but it's *more* true for people than for say spheres of lead falling in vacuum.
Particularily with problems that are uncommon or rare it's a huge problem to get enough of a sample-size that there's still any sample left after you correct for the obvious and/or known factors.
Take SIDS in Europe for example. Incidence is 1:5000 or thereabout. We know that smoking increases the risk quite a lot, so any study that wants to do research on *other* factors needs to factor for smoking or non-smoking.
We also know that low birth-weigtth, young mothers, insufficient pre-natal care and certain sleeping-conditions have an effect. Once you split for all of these, you'd need a gargantuan start-population to have anything left at all.
1:5000 children will die of SIDS. 1:3 children have parents that smoke. (2:3 of the SIDS-dead) 3% of all children have low birth-weigth. etc.
If you wanted to do research on other theories, say the theory that gases given of from certain foam matresses play a role, you'd need to have initial data for literally millions of children to have even a *hope* of correctly canceling all those other effects an narrow in on the influence (or noninfluence) of mattresses.
As if this wasn't bad enough:
PArticipation in medical experiments is generally voluntarily, and some people refuse or drop out during the experiment or whatever. But and that's the tricky part -- the ones dropping out are *NOT* a random part of the population, but rather a certain type of people, so this f*cks up your data too and needs to be corrected, as good as you manage.
This stuff is hard. Very hard.
If I create the trademark "Quropiwla", which is a freely invented word, then I'm going to have a fair chanse demonstrating that anyone using that word does it in the intention of creating confusion with my product -- afterall, what other reason could there be ? It's unlikely someone would end up with that word by chance, and the word doesn't have any meaning as such.
If, on the other hand I register "Enjoyable" in the market for computer-input-devices (say I market joysticks) then it's not that obvious that others using this word does so in order to create confusion with my product. I'd probably still be able to stop other "Enjoyable" joysticks, but I doubt I'd have a case against say the "Enjoyable" computer-monitor.
It's likely the poster meant it as you explain, but I don't exactly think you are the rigth one to talk about being rude or having an attitude.
Unless, offcourse, where you come from "WTF" and thelike is considered polite.
Besides, I never even set foot in any American school, WTF is up with people assuming everyone and everything are American ?
If a rotor on a moving craft moves quickly enough that the tip-speed is equal to the speed of the craft, then the tip of the rotor will stand still, relative to the surrounding air on its way backwards.
But all other parts of the rotor move *slower* than the tip, so no part of the rotor will move backwards relative to the surrounding air. That's just bullshit.
I dunno, how much does MS have in the bank ? More than $170 billion ? I kinda doubt it.
So the question is, how many of the ones that buy a new server *after* these cards become available on the market will install a kernel older than 2.6.10?
Perhaps a few, but probably *not* the ones for which performance is important, nor the majority which simply takes whatever is default in their distro.
Fedora and Mandriva both have newer default kernels (2.6.11 I think), and I guess even Debian uses 2.6 now ?
With newer Linux-kernels this is quite simply not the case.
To avoid the torrent of interupts from a fast nic the Linux-kernel detects that the card gets packets so often that essentially there's "always" one or more packets waiting in the cards buffers.
It responds to this condition by disabling interupts for the card in question and switch to polling it regularily.
Normally polling is inefficient, because it amounts to asking over and over again "got anything for me now?", where in most situations the answer is "no" 99.99% of the time, which makes it a waste of resources to ask in the first place.
This changes when the answer is "yes" basically all the time. There's no need for the network card to tell the cpu over and over and over "I got a packet for you", instead the cpu collects packets regularily.
It's sorta like having your lawyer receive legal letters for you.
If you get very few, it'd be a waste for you to drop by him every day and ask if he's gotten any for you (polling) most of the days you'd be making the trip for naught. In this situation interupts (i.e. having your lawyer call you and inform you on the rare occasions when a letter *does* arrive) makes more sense.
But if your letter-flow increases to the point where there's normally 3-5 letters every day and it's rare that no letter arrives, then it no longer makes sense that the lawyer calls you every day to tell you you got letters (interupts), you already assumed that. In this scenario it makes more sense for you to come by regularily and pick up letters without being prompted to do so. (polling)
Thus, the flow of interupts from a Gb-nic being flooded with 100byte small-packets (say a loaded dns-server) is not 1 million interupts every second -- it is zero interupts.
(allthough what you write is correct for kernels older than 2.6.10 and for less clever OSes.)
Sure, your "fix" sort of helps. A little. For those using english. Assuming they're using the "correct" font (some glyphs look similar or not depending on the fonts used)
"No good reason" to use international domain-names in a trustworthy domain name ? Isn't that the same as saying there's no good reason to use them at all ?
The *obvious* use of international domain names is for companies to be able to use the domain name corresponding to their actual name.
Skoda, Møreforskning, Brønnøysundregisteret, Ålesund, Autohaus-Lübeck
If you can't do that (reliably) there's no (or very little) point in having international domain names at all. Your "fix" thus isn't a "fix" unless you consider: remove all traces of idn to be a "fix".