At least for criminal matters, they need to have timestamps when they request what customer had what IP address. (When using it in court, the timestamps had better be accurate and the search warrant better have produced corroborating evidence.) Since the ISP and investigators know full well how DHCP works, they almost always give out the correct information for the IP (since they keep timestamped logs).
Weirdly, I am simultaneously: a) offended that they're applying such arbitrary restrictions to teachers -- it seems, technologically, stupid, and I've had teachers that are friends b) pleased, since facebook is a pretty popular venue for creepy guys, and c) surprised that teachers associating with students on facebook is a big problem, since it seems extremely unprofessional to me.
I have friends that teach and are Facebook friends with their *former* students, but association outside the classroom really should be conducted in a professional manner.
No, in science, you modify your model and conclusions based on changing evidence. The difference here is that you're holding your conclusion constant and changing the reason you claim it's true every time your reason is found to be untrue.
It's stupid because it's unquantified. There are seven planets that aren't Earth. Temperatures of reasonably complicated systems are rarely constant. So over any particular span of time, each of those seven planets will have either heated up or cooled down. Without information about how much and why, the fact that the Earth's temperature deviated in the same direction is meaningless.
If I turn my oven on in the morning, you could equally note that the outdoors are heating up (from about 40 F to 65 F over the course of the day) and the oven is also heating up (70 F to 350 F over some minutes).
Oh, no, your physics calculations totally depend on frame rate. Just not resolution.
You can actually cheat and gate things like physics and AI at around 60 fps and then let the animations run faster than that, which can be useful especially if your AI is expensive, but that's less common.
Games generally do use what is essentially floating-point time, but update every frame, because modern games tend to have distinctly non-constant frame rates.
I agree 2000 is pretty small, but a lot of the words at the end of your list aren't even words and many others are not at all easy to remember. Kms, lta, meps, mhz, mics, nsc, owd, pac?
You actually want to use not short words but extremely common words, since they are more likely to be easily remembered. 2000 was chosen because of the XKCD about this password generation method, but you could probably use a factor of two more and still have common-enough words.
Remembering is, of course, the hard part with passwords. For non-remembered passwords, I like to make the character space as large as possible and the password length fairly large (~12 char), sometimes omitting visually-similar characters.
Well, GPU cracking is something like 500 million hashes / sec = 2^29 hashes/sec. Four words out of a 2k-word dictionary (which is small), selected randomly, is a space of 2^44 passwords. That's about 9 GPU-hours, which is not good. Adding a fifth word increases this to roughly 2 GPU-years (a factor of 2^11). Adding numbers in between the four words increases the password space by about 2^5, which is something (~300 GPU-hours) but is not really substantial. (A sixth work makes it 4000ish GPU-years, which is starting to get really cost prohibitive.)
More effective, really, is for people storing passwords to increase the cost of computing hashes. If you use something like HMAC, both cracking time and password verification time scale linearly in the number of rounds. Client-side, this is easy. Well-designed modern encryption software, for example, uses enough rounds in password-based key derivation that it takes on the order of a second to compute. That's roughly a million rounds, so password cracking against a 4-word password at 500 Mhashes/sec increases from 9 GPU-hours to 1000 GPU-years. Server-side, password verification is more expensive, but even using thousands of rounds of SHA1 over one round of MD5 is a huge security increase.
Unfortunately, the end user has little control (or even knowledge) of how passwords are stored server-side.
The more effective route is to have public funding agencies demand this. If a university demands it, then it can substantially limit the capacity of a researcher to publish effectively with little upside. If the funding agency demands it, then the researcher has the perfect excuse -- nobody is going to tell the guy with the money "no thanks" just because he wants you to publish in a lower-impact free journal.
Ideally journals themselves would be replaced with a decentralized Web based system where anyone can publish and peers can freely review all the articles.
Many researchers post preprints on arXiv and/or on their personal web sites. There's decentralized free peer review, too, but only if you show up to colloquia.
I doubt it's helpful, but Apple's disk image encryption (really, the disk images one layer below the encryption) supports that. Sparsebundles automatically work that way; you can also manually create split images of a predefined size (though perhaps only on the command line).
I think if your gear could take 50g, which is respectably high, you'd need a 100 km track to accelerate to 11 kps, and that escape velocity doesn't take air resistance into account.
If you hadn't bought music on the iPhone and hadn't linked it with a copy of iTunes yes, what iTunes stuff was there on the phone to erase and how did it get there?
Since you were working with a copy of iTunes that had no Apple ID, if she bought albums via iTunes for her phone, it wouldn't have been a problem. You associate the existing copy of iTunes with the same Apple ID as used on the iPhone, connect the iPhone, and Transfer Purchased Content. (You could also just find and download all purchased content in the iTunes Store without using the iPhone at all, once iTunes is linked to the Apple ID, but that would require redownloading.) Now that your copy of iTunes has all of the purchased content from the iPhone on it, you can sync your iTunes library to your iPhone without losing data.
"Go boom" is just an inaccurate way of describing a particle-antiparticle interaction. Particle interactions generally involve some particles going in and some other particles, or particles in different states, coming out. If a particle and it's antiparticle interact, neither of them are products of the interaction. (Often the product is just photons.)
It's in theory more impressive if you have bulk antimatter, since matter holds a ton of energy. Individual particles, though, have relatively little energy, so their annihilation is not so spectacular.
Some particles are already their own antiparticle. Photons, for example.
Also, note that there cannot exist matter that "cannot turn into" energy. All matter, by definition, has mass. Mass and energy are the same thing. So all particles contain a nonzero amount of energy. There are no known "noble" particles: that is, one that cannot interact with another particle and thereby transfer its energy.
Antiparticles are not just particles with opposite electric charge. They're not magically, fundamentally different, but other particle properties are negated as well.
If it was just electric charge, then neutral-charge particles would have no antiparticle. But they do -- for example, the neutrinos (and antineutrinos).
Depending on the type of pen and the pressure used, that method may not be sensitive enough for good reconstruction of the data. However, since it's a destructive technique, if you try it first and it fails, you've ruined your ability to try any other techniques.
This is really true of any pair of countries. The only reason to host data and servers in the US is if it's (much) cheaper or if that data needs to be highly available to your customers in the US. Otherwise, the legal and practical implications of storing your data in a country other than your own make such a decision crazy for businesses.
Slashdot Mirror
At least for criminal matters, they need to have timestamps when they request what customer had what IP address. (When using it in court, the timestamps had better be accurate and the search warrant better have produced corroborating evidence.) Since the ISP and investigators know full well how DHCP works, they almost always give out the correct information for the IP (since they keep timestamped logs).
Weirdly, I am simultaneously:
a) offended that they're applying such arbitrary restrictions to teachers -- it seems, technologically, stupid, and I've had teachers that are friends
b) pleased, since facebook is a pretty popular venue for creepy guys, and
c) surprised that teachers associating with students on facebook is a big problem, since it seems extremely unprofessional to me.
I have friends that teach and are Facebook friends with their *former* students, but association outside the classroom really should be conducted in a professional manner.
No, in science, you modify your model and conclusions based on changing evidence. The difference here is that you're holding your conclusion constant and changing the reason you claim it's true every time your reason is found to be untrue.
It's stupid because it's unquantified. There are seven planets that aren't Earth. Temperatures of reasonably complicated systems are rarely constant. So over any particular span of time, each of those seven planets will have either heated up or cooled down. Without information about how much and why, the fact that the Earth's temperature deviated in the same direction is meaningless.
If I turn my oven on in the morning, you could equally note that the outdoors are heating up (from about 40 F to 65 F over the course of the day) and the oven is also heating up (70 F to 350 F over some minutes).
Neither of your examples are remotely similar to modern games.
Oh, no, your physics calculations totally depend on frame rate. Just not resolution.
You can actually cheat and gate things like physics and AI at around 60 fps and then let the animations run faster than that, which can be useful especially if your AI is expensive, but that's less common.
Games generally do use what is essentially floating-point time, but update every frame, because modern games tend to have distinctly non-constant frame rates.
Perhaps part of your problem is making the developers do the art.
You know, calculating collisions is independent of resolution.
I agree 2000 is pretty small, but a lot of the words at the end of your list aren't even words and many others are not at all easy to remember. Kms, lta, meps, mhz, mics, nsc, owd, pac?
You actually want to use not short words but extremely common words, since they are more likely to be easily remembered. 2000 was chosen because of the XKCD about this password generation method, but you could probably use a factor of two more and still have common-enough words.
Remembering is, of course, the hard part with passwords. For non-remembered passwords, I like to make the character space as large as possible and the password length fairly large (~12 char), sometimes omitting visually-similar characters.
Well, GPU cracking is something like 500 million hashes / sec = 2^29 hashes/sec. Four words out of a 2k-word dictionary (which is small), selected randomly, is a space of 2^44 passwords. That's about 9 GPU-hours, which is not good. Adding a fifth word increases this to roughly 2 GPU-years (a factor of 2^11). Adding numbers in between the four words increases the password space by about 2^5, which is something (~300 GPU-hours) but is not really substantial. (A sixth work makes it 4000ish GPU-years, which is starting to get really cost prohibitive.)
More effective, really, is for people storing passwords to increase the cost of computing hashes. If you use something like HMAC, both cracking time and password verification time scale linearly in the number of rounds. Client-side, this is easy. Well-designed modern encryption software, for example, uses enough rounds in password-based key derivation that it takes on the order of a second to compute. That's roughly a million rounds, so password cracking against a 4-word password at 500 Mhashes/sec increases from 9 GPU-hours to 1000 GPU-years. Server-side, password verification is more expensive, but even using thousands of rounds of SHA1 over one round of MD5 is a huge security increase.
Unfortunately, the end user has little control (or even knowledge) of how passwords are stored server-side.
The more effective route is to have public funding agencies demand this. If a university demands it, then it can substantially limit the capacity of a researcher to publish effectively with little upside. If the funding agency demands it, then the researcher has the perfect excuse -- nobody is going to tell the guy with the money "no thanks" just because he wants you to publish in a lower-impact free journal.
Ideally journals themselves would be replaced with a decentralized Web based system where anyone can publish and peers can freely review all the articles.
Many researchers post preprints on arXiv and/or on their personal web sites. There's decentralized free peer review, too, but only if you show up to colloquia.
I doubt it's helpful, but Apple's disk image encryption (really, the disk images one layer below the encryption) supports that. Sparsebundles automatically work that way; you can also manually create split images of a predefined size (though perhaps only on the command line).
Radiative heat is actually just photons.
Photons have mass and carry momentum, but don't have rest mass.
Neither of which are heat.
If you wait a few seconds, your watts will turn into joules.
I think if your gear could take 50g, which is respectably high, you'd need a 100 km track to accelerate to 11 kps, and that escape velocity doesn't take air resistance into account.
If you hadn't bought music on the iPhone and hadn't linked it with a copy of iTunes yes, what iTunes stuff was there on the phone to erase and how did it get there?
Since you were working with a copy of iTunes that had no Apple ID, if she bought albums via iTunes for her phone, it wouldn't have been a problem. You associate the existing copy of iTunes with the same Apple ID as used on the iPhone, connect the iPhone, and Transfer Purchased Content. (You could also just find and download all purchased content in the iTunes Store without using the iPhone at all, once iTunes is linked to the Apple ID, but that would require redownloading.) Now that your copy of iTunes has all of the purchased content from the iPhone on it, you can sync your iTunes library to your iPhone without losing data.
"Go boom" is just an inaccurate way of describing a particle-antiparticle interaction. Particle interactions generally involve some particles going in and some other particles, or particles in different states, coming out. If a particle and it's antiparticle interact, neither of them are products of the interaction. (Often the product is just photons.)
It's in theory more impressive if you have bulk antimatter, since matter holds a ton of energy. Individual particles, though, have relatively little energy, so their annihilation is not so spectacular.
Some particles are already their own antiparticle. Photons, for example.
Also, note that there cannot exist matter that "cannot turn into" energy. All matter, by definition, has mass. Mass and energy are the same thing. So all particles contain a nonzero amount of energy. There are no known "noble" particles: that is, one that cannot interact with another particle and thereby transfer its energy.
Neutrinos are just an example, but yes.
Antiparticles are not just particles with opposite electric charge. They're not magically, fundamentally different, but other particle properties are negated as well.
If it was just electric charge, then neutral-charge particles would have no antiparticle. But they do -- for example, the neutrinos (and antineutrinos).
X-ray fluorescence spectroscopy is a lot more fun.
Depending on the type of pen and the pressure used, that method may not be sensitive enough for good reconstruction of the data. However, since it's a destructive technique, if you try it first and it fails, you've ruined your ability to try any other techniques.
This is really true of any pair of countries. The only reason to host data and servers in the US is if it's (much) cheaper or if that data needs to be highly available to your customers in the US. Otherwise, the legal and practical implications of storing your data in a country other than your own make such a decision crazy for businesses.