Really, the problem here is overselling capacity in a batshit crazy fashion. You can oversell capacity, and do it sanely. Such as actually metering actual network utilization over time, and oversell by perahps 10 to 20%. Instead, these carriers are pathologically allergic to improving their infrastructures, and pathologically oversell their capacity, to the point where they think using more than 100mb in a month is "heavy use". News flash: if you have lots of apps installed on your phone, simply enabling the autoupdater will push you over that pathetically small limit.
What is taught in school (communications engineering), is that you oversell as much as you can get away with, because that's where the profit (or ability to compete with price) is.
There is one unique problem with mobile networks: when the operator sells you a phone, they have little clue about where exactly you are going to use that phone. So at sale time, it is not possible to determine if this particular handset is going to make their network laggy. The only option is to monitor the network and react to congestion afterwards.
When congestion hits, the best option (for average user experience) would be to have some sort of fair share rate limiting to kick in. Users who are trying to stream something will quickly get the clue and stop trying (who wants to watch movies in 2sec play - 5sec buffer loop). And naturally, the operator should look for ways to improve their network in often congested areas.
Because there are two dimensions to mobile network congestion (time and place), I think data caps are completely insane way to try to solve the issue. If you use all your data being the only customer currently being served by the cell, no one else gets disrupted. Well implemented rate limits and QoS would make much more sense as congestion control.
Btw. I'm paying ~15€/mo for uncapped full speed mobile data (ToS forbids p2p but doesn't seem to be actually enforced). I've used about 2GB total in the last 6 months. Turns out I spend most of my time near a friendly WLAN.
Adding to parent, they could probably have made their point by nuking much less densely inhabited areas. Instead they decided to go for maximum civilian casualties.
[...] ban Twitter [...] cell phones? Better ban those [...] outlaw cars and remove the roads [...] entire government division to enforce and strictly regulate corrective vision dealers [...]
That's an awful lot of work. Wouldn't it be much easier to just ban the terrorists?
Also, in N900, the telephone functionality UI is just an "app" itself, so no special moves are required to get out of the phone screen. Just your normal every-day app switching.
If you want to know who is having an affair with whom just look for correlation in holidays and sickleave, you don't need to abuse the IT systems.
Apparently our workplace is a rather polygamous community, most people are having their vacations simultaneously! Even sick leaves seem to often happen in clusters of several people, and that's not even the same groups all the time! This certainly does explain some things.
You have to be trolling. There is no way you believe that you can prove two numbers are non-random.
If he believes (as he seems to imply) that the set from which the two numbers are from is the set of positive integers (which is infinite), and that the random samples are uncorrelated, then it indeed seems very likely that those numbers are not random. Then there is also the subtle difference between "prove logically" and "prove statistically" which both often shorten to just "prove".
This is a simplification that is not always true. What you want is a sensor size that is properly matched to the lens. In SLRs, cheaper bodies have sensors that are smaller than the total image projected onto the focal plane, so that light through the lens is wasted. If the sensor size is properly matched to the lens, you will get the best quality.
So, yes, if you buy an SLR, probably sensor size matters, but what is more important is matching the sensor size to the lens system.
in a sense, you are correct. However, entry-level DSLRs typically come with a glass that's meant for the sensor size of the body it's sold with. And even lenses that are meant for larger image area typically exhibit their best properties near the center of the image, so one could say that the mid-size sensor is right in the sweet spot of the projection.
+1.
Entry-level DSLRs are bulkier and a bit more expensive than point-and-shoots and ultrazooms, but the difference in price is not that great. In return for the investment you will get something the small stuff lacks in: larger image sensor. In image quality, size does matter.
Why not? I can think of lots of people, most of them would be scientists though, rather than authors but I see no reason why people shouldn't be allowed amass vast wealth if they do something to better society.
If they keep it circulating, it doesn't matter. It helps to keep things going.
On the other hand, if they're allowed to sit on it, that's bad.
The algorithm for choosing a port at random is security by obscurity.
If the algorithm is keyed, publicly known, and known to provide good security, then it's not security through obscurity. At least according to some some other comments here.
Of course, the result space (2^16) is too small to provide actual security.
Slashdot Mirror
This has been bothering me lately. If we evolved from monkeys, why are they so stupid?
When jumping from one tree to another, bigger (and heavier) brain is not an advantage in selection.
Really, the problem here is overselling capacity in a batshit crazy fashion. You can oversell capacity, and do it sanely. Such as actually metering actual network utilization over time, and oversell by perahps 10 to 20%. Instead, these carriers are pathologically allergic to improving their infrastructures, and pathologically oversell their capacity, to the point where they think using more than 100mb in a month is "heavy use". News flash: if you have lots of apps installed on your phone, simply enabling the autoupdater will push you over that pathetically small limit.
What is taught in school (communications engineering), is that you oversell as much as you can get away with, because that's where the profit (or ability to compete with price) is.
There is one unique problem with mobile networks: when the operator sells you a phone, they have little clue about where exactly you are going to use that phone. So at sale time, it is not possible to determine if this particular handset is going to make their network laggy. The only option is to monitor the network and react to congestion afterwards.
When congestion hits, the best option (for average user experience) would be to have some sort of fair share rate limiting to kick in. Users who are trying to stream something will quickly get the clue and stop trying (who wants to watch movies in 2sec play - 5sec buffer loop). And naturally, the operator should look for ways to improve their network in often congested areas.
Because there are two dimensions to mobile network congestion (time and place), I think data caps are completely insane way to try to solve the issue. If you use all your data being the only customer currently being served by the cell, no one else gets disrupted. Well implemented rate limits and QoS would make much more sense as congestion control.
Btw. I'm paying ~15€/mo for uncapped full speed mobile data (ToS forbids p2p but doesn't seem to be actually enforced). I've used about 2GB total in the last 6 months. Turns out I spend most of my time near a friendly WLAN.
Adding to parent, they could probably have made their point by nuking much less densely inhabited areas. Instead they decided to go for maximum civilian casualties.
[...] ban Twitter [...] cell phones? Better ban those [...] outlaw cars and remove the roads [...] entire government division to enforce and strictly regulate corrective vision dealers [...]
That's an awful lot of work. Wouldn't it be much easier to just ban the terrorists?
Or the fundamentalist Christians, for that matter.
Religion supports terrorism, it needs to be banned.
Otherwise, the real story here is how Namibia came to be floating over South Africa.
Yeah, saw that movie, I think it was called District 9.
Also, in N900, the telephone functionality UI is just an "app" itself, so no special moves are required to get out of the phone screen. Just your normal every-day app switching.
Point the finger where the finger is due.
Agreed.
Why isn't EU proposing to ban use, sales and manufacturing of all surveillance tech in it's own member countries? There's something to think about.
I though there was some commission to prevent mergers that cause harm to the general public?
Radical tube rides, dude!
If you want to know who is having an affair with whom just look for correlation in holidays and sickleave, you don't need to abuse the IT systems.
Apparently our workplace is a rather polygamous community, most people are having their vacations simultaneously! Even sick leaves seem to often happen in clusters of several people, and that's not even the same groups all the time! This certainly does explain some things.
You have to be trolling. There is no way you believe that you can prove two numbers are non-random.
If he believes (as he seems to imply) that the set from which the two numbers are from is the set of positive integers (which is infinite), and that the random samples are uncorrelated, then it indeed seems very likely that those numbers are not random. Then there is also the subtle difference between "prove logically" and "prove statistically" which both often shorten to just "prove".
Then again, the set could as well be {3, 7}.
Server Apache/2.2.3 (Red Hat) X-Powered-By PHP/5.2.17
'Nuff said.
Still on PHP 5.2? They could hire an admin to do updates.
... 25 years in fighting off all the complaints from various parties.
5 years in actual construction work.
This is a simplification that is not always true. What you want is a sensor size that is properly matched to the lens. In SLRs, cheaper bodies have sensors that are smaller than the total image projected onto the focal plane, so that light through the lens is wasted. If the sensor size is properly matched to the lens, you will get the best quality. So, yes, if you buy an SLR, probably sensor size matters, but what is more important is matching the sensor size to the lens system.
in a sense, you are correct. However, entry-level DSLRs typically come with a glass that's meant for the sensor size of the body it's sold with. And even lenses that are meant for larger image area typically exhibit their best properties near the center of the image, so one could say that the mid-size sensor is right in the sweet spot of the projection.
+1. Entry-level DSLRs are bulkier and a bit more expensive than point-and-shoots and ultrazooms, but the difference in price is not that great. In return for the investment you will get something the small stuff lacks in: larger image sensor. In image quality, size does matter.
Duh.. mispasted the link. http://abcnews.go.com/US/story?id=3642673
Funny you should mention that, according to this there's no woman oppresion either.
- How long before someone turns the blocking of "lesbian" and "gay" into a human rights issue? Especially "gay pride"
In muslim countries, they have no gays.
That's how the French say 80.
Flying cars are already a reality. They are just expensive and inefficient.
Any car can fly. Flight distance and safe landing are the real problems...
Why not? I can think of lots of people, most of them would be scientists though, rather than authors but I see no reason why people shouldn't be allowed amass vast wealth if they do something to better society.
If they keep it circulating, it doesn't matter. It helps to keep things going.
On the other hand, if they're allowed to sit on it, that's bad.
So the question is: where the bug actually is?
In the motherboard BIOS. But for some reason, no one here is demanding an update from mobo vendors... why?
The algorithm for choosing a port at random is security by obscurity.
If the algorithm is keyed, publicly known, and known to provide good security, then it's not security through obscurity. At least according to some some other comments here.
Of course, the result space (2^16) is too small to provide actual security.
Many crytpo schemes are provably secure.
Actually, only one: the one-time pad.