Battersea looks like an ideal location though. It's on the river, so has a good source of water for cooling, and it's surrounded by large electricity consumers.
The flow rate's not great though; better to build a very large plant further off on the coast (e.g., Bradwell and Sizewell) and use very-high-voltage transmission lines to bring it in. Which is exactly what they're going to do, funnily enough...
If you're a great programmer already, why even bother with the CS degree?
There's only one good answer to that question, you know: because you truly want to understand computers, their deep nature, their history, their use, everything. CS isn't about getting a job, it's about answering a deep calling. I suspect that that's the way it should be for any degree; I hope it is, because if so, there are people out there who will be deriving some of the best possible satisfaction ever from taking those courses.
If college is just a step towards a job to you, why not go to trade school instead? Leave the university places free for those who really want to learn.
Personally, I feel that logic should be taught in elementary school. It's a joke that we expect our grade school students to write argumentative essays without understanding the nuts and bolts of the argument. We should also teach the fundamentals of philosophy in high school. Most people go to college knowing who Socrates, Plato, and Aristotle are, but they know nothing of their ideas.
Nor should they be taught all that those three thought, as we have in first-order logic a far stronger foundation for logic than syllogisms. In particular, it's far easier to avoid making mistakes with FOL (and its little brother, boolean logic) precisely because it is far less ambiguous; you have to write what you mean and in the process you understand what you mean far better. Because FOL is simple in presentation (if not in its consequences!) it's reasonable to consider teaching it to schoolchildren. Syllogistic logic is inherently trickier to use.
From personal experience, I can tell him English is equally important. I'm not overly qualified when it comes to CS but I work on computers. You're not always communicating with other tech people and oftentimes you deal with people who don't speak English natively. Without writing skills you're bound to have communication problems. If your boss needs a tech manual along with that program you wrote, it better not have any grammatical errors.
Agreed. As a programmer, software engineer and computer scientist, I write far in English than in any programming language. Communication with other people is and will remain vital. The OP had better be prepared to read and write masses of stuff, every day for the rest of their working life... (I don't mind myself; I like reading and don't mind writing too much...)
It has been known for some time that a botnet’s combined computing power could be used for a number of nefarious purposes. We can now add Bitcoin mining to that list.
So that is what all those annoying Flash things in webpages are doing when they swallow all my CPU power!
Parent probably means countries where the bedrock isn't close to the surface. In some places, it's more than a mile deep, which isn't going to be very practical.
Most countries have bedrock closer to the surface than that, even if not everywhere. Moreover, mining to more than a mile down isn't too hard, especially if you're not digging through a coal seam (when you would have gas problems). The main issue with deep mines is usually just water ingress, but not all sites have that problem. For example, Boulby (a salt mine) is nearly a mile deep. The only reason we don't normally go down that far is because it's expensive and what we're after is typically closer to the surface.
Except that marijuana smoke does not contain any Tobacco-specific nitrosamines, and the evidence so far is that marijuana smokers are not more likely to develop cancer than non-smokers. That aside, smoking is not the only means by which marijuana is consumed, and non-smoking methods of use appear to have no permanent effects (as opposed to non-smoking methods of using tobacco, which still increase the risk of cancer).
Look, it's hard to believe that putting smoke of any kind in your lungs is actively good for you, whether it comes from tobacco, or marijuana, or even goddamn soft furnishings. As such, if marijuana is legalized then it would be reasonable to regulate its consumption by smoking in the same way as consumption of tobacco through the same route is regulated. Moreover, it has some major effects that are similar to alcohol (slows reaction times, impacts judgement) so there would also be good reason to use some of the same sorts of regulations that are used with alcohol (e.g., don't drive when intoxicated, however you got into that state). These sorts of things cover the key public harms that need to be tackled upon legalization, and the other issues are either none of the governments business anyway (people have a right to harm their own health, even if it is stupid for them to do so) or are a direct consequence of prohibition (the enormous majority of criminality associated with substance use).
In short: legalize and regulate through adaptation of existing rules for other (legal) drugs. You Know It Makes Sense.
The best would be the mode, i.e., the speed that customers are most likely to see. The median would also be good. They're also not that easy to scam (except by outright lying, which it's a bit too easy to be caught in).
The median household income in the US is $44k. The mean household income in the US is $60k. Where did you get $31k?
It could have been the modal income; the mode is usually the lowest of the three major average types, for the kinds of distribution seen in things like incomes (where there's a long upper tail but much more bunching up at the bottom end).
Also, the GGP is talking about personal income whereas the GP is talking about household income. The majority of households will have more than one income-earner so things are still consistent.
The median household income in the US is $44k. The mean household income in the US is $60k. Where did you get $31k?
It could have been the modal income; the mode is usually the lowest of the three major average types, for the kinds of distribution seen in things like incomes (where there's a long upper tail but much more bunching up at the bottom end).
Speaking from a statistical PoV. I've no idea what the values actually are.:-)
...somehow I don't think that a multi-agency-operation can be executed within 1 day
If you show your ass to authorities on six different continents, it goes without saying they're going to feel a lot more generous about cooperating in capturing you.
But it will still take more than a day for them to agree to arrest you. Don't underestimate the need for the management to have a good long meeting first, and the more high-profile agencies involved, the more that becomes necessary. (It's a primate thing, rather like baboons showing off the colour of their genitals.)
I don't think that law would actually hold up in court against plausible deniability encrypted disk images. I think the judge would want proof that a hidden volume actually did exist.
I'd hate to try to use such a defence in a trial. It'd be far more likely that the jury (or triumvirate of magistrates, in a lower-level court) would decide that you were pissing around with them and state that you were guilty, especially if the police provided evidence that you were communicating in a way that indicates that you were involved in hacking. Remember, conviction requires that it be proved beyond reasonable doubt, not beyond all doubt.
This is the UK. Should he be someone from lulzsec and if they have a decent amount of evidence to prove he was a main player I'd say he'll get 2-3 years max and likely out in 12-18 months for good behaviour
At that sentence length, no. Maximum 1/3 off for good behaviour once the sentence is over 2 years long.
You don't need SELinux nor Vsys, you can do that using POSIX file permission. I have a 'firefox' user which can only read write to a 'Downloads' directory, and nothing else.
Now try scaling that up to having two real users on the same system. Or 5 users. Or 20.
Which is why the UK is known for creating all these cloud based services? I must have missed the memo, but nearly all the ones I can think of are from the US. Admittedly, I hear about more of the ones from here because of sample biasing, but I doubt very much that it's that far from the truth.
That's actually mostly sample biasing. UK providers tend to be focused towards selling into either the UK or EU markets, especially dealing with cases where going to the US is utterly not an option, e.g., you'll see very few EU-based firms locating data that contains anything even vaguely personally-identifiable in the US, and that's entirely driven by differences in law, and I've worked with firms that were putting things in the cloud which were even more restricted than that, but again, going to the US wasn't an option, not without proving that Official Secrets were not being exported (we could have proved it for sure, but it was several years of hassle we didn't want).
The UK is also not a particularly attractive place to site a major datacenter: power isn't particularly cheap, and land is (still) very expensive. This encourages UK firms to outsource low-level cloud provision where possible...
Every piece of automation that replaces a person puts that person out of work at that job.
The real question is whether it is better for that person to not be doing that job in the first place. Perhaps they'd be better off overall doing something else. If that's so, then that old job is better automated. (OTOH, just doing them out of a job and leaving them on the dole is not better overall; joblessness is the source of all sorts of social ills. Make-work is better than no-work, even if nowhere near as good as real work.)
Those that do not understand how Public Key Crypto works should not use it.
In other news, gweihir has announced that he will no longer be accessing any website via HTTPS.
(The number of people who understand the whole of a public key crypto system and deployment is vanishingly small. The underlying math is difficult. The programming is easy to make errors in. The way to use it, not all that obvious either going by the massive quantities of misinformation I see here and elsewhere on the 'net. Public key crypto is only practical to use if you don't understand it all; fortunately, there are useful abstractions for most of it that are accessible. Now, if only the firmware makers grokked even that little bit...)
British Rail (which is one of the worst networks in Europe) runs goods trains FASTER than most passenger trains.
It depends on the metric. UK railways tend to run with thinner margins than many other European railways, which means that they've got more opportunity for things to go wrong. The problem is just that UK traffic levels are very high for the amount of number of tracks on any particular route, so failures are much more likely to cause cascading problems. In particular, a lot of routes only have one or two tracks and management hates the thought of investing to fix this.
Look at what has worked and why, look at what hasn't worked and why, but don't look at what hasn't worked and assume that this is a universal constant.
That's true of reference counting (shared_ptr), not unique_ptr/scoped_ptr, std::array, vector or similar which compile to equivalent C code.
But then you've got objects of the lifespan implied by those entities. If you want to have objects that don't have a tight owning scope or container, then you've got to either put the free/delete in the right place (hard!) or use reference counting, or use garbage collection. You can wrap up lots of stuff around that, but ultimately that's the type of technique you need. That in turn means you've got to pay the cost of the technique. C++'s advantage is that it gives you the flexibility to use a variety of techniques as appropriate. C++'s disadvantage is that you have to select the right technique and it's not guaranteed to be nice if you get it wrong. (Code that runs in a managed environment is more expensive, but much less destructive when things go awry. It's a trade-off.)
Except it is legal in the UK, and has been for decades. Assisting is not legal, mainly to prevent the unscrupulous from manipulating other people into killing themselves. (Some people are utter scum, alas.) It's been stated clearly that while each case of alleged assisting a suicide will be taken on its merits, when it's clear that it's a genuine choice of the suicidee that's not been forced upon them in some way and where there hasn't been too much assistance, there won't be prosecution. That's a wooly position, but just about anything else makes it too easy for the callous (especially if they're family members) to abuse in some way.
Suicide is a terrible thing. It's so unfortunate that it is sometimes the best option. As a society, we should try to ensure that there are as many positive, constructive alternatives to killing oneself as possible. (But if you do decide to go, don't do it by jumping in front of a train or other vehicle; that's very inconsiderate!)
Slashdot Mirror
Ruby is dying fast.
That's some of the most robust "dying fast" I've seen in a while. How many decades do you think we'll need to wait before ordering the hearse?
All you have to do is send a blank check to this Nigerian prince...
When you do send it, don't forget to sign it "Bernie Madoff".
Actually, wind and geothermal should play a big part of UK (and USA's) energy future.
The UK has good geothermal sites? That's odd, since it's a long way from significant thermal sources (no active volcanoes here!)
Battersea looks like an ideal location though. It's on the river, so has a good source of water for cooling, and it's surrounded by large electricity consumers.
The flow rate's not great though; better to build a very large plant further off on the coast (e.g., Bradwell and Sizewell) and use very-high-voltage transmission lines to bring it in. Which is exactly what they're going to do, funnily enough...
If you're a great programmer already, why even bother with the CS degree?
There's only one good answer to that question, you know: because you truly want to understand computers, their deep nature, their history, their use, everything. CS isn't about getting a job, it's about answering a deep calling. I suspect that that's the way it should be for any degree; I hope it is, because if so, there are people out there who will be deriving some of the best possible satisfaction ever from taking those courses.
If college is just a step towards a job to you, why not go to trade school instead? Leave the university places free for those who really want to learn.
Personally, I feel that logic should be taught in elementary school. It's a joke that we expect our grade school students to write argumentative essays without understanding the nuts and bolts of the argument. We should also teach the fundamentals of philosophy in high school. Most people go to college knowing who Socrates, Plato, and Aristotle are, but they know nothing of their ideas.
Nor should they be taught all that those three thought, as we have in first-order logic a far stronger foundation for logic than syllogisms. In particular, it's far easier to avoid making mistakes with FOL (and its little brother, boolean logic) precisely because it is far less ambiguous; you have to write what you mean and in the process you understand what you mean far better. Because FOL is simple in presentation (if not in its consequences!) it's reasonable to consider teaching it to schoolchildren. Syllogistic logic is inherently trickier to use.
From personal experience, I can tell him English is equally important. I'm not overly qualified when it comes to CS but I work on computers. You're not always communicating with other tech people and oftentimes you deal with people who don't speak English natively. Without writing skills you're bound to have communication problems. If your boss needs a tech manual along with that program you wrote, it better not have any grammatical errors.
Agreed. As a programmer, software engineer and computer scientist, I write far in English than in any programming language. Communication with other people is and will remain vital. The OP had better be prepared to read and write masses of stuff, every day for the rest of their working life... (I don't mind myself; I like reading and don't mind writing too much...)
17th June, Symantec's blog:
It has been known for some time that a botnet’s combined computing power could be used for a number of nefarious purposes. We can now add Bitcoin mining to that list.
So that is what all those annoying Flash things in webpages are doing when they swallow all my CPU power!
Parent probably means countries where the bedrock isn't close to the surface. In some places, it's more than a mile deep, which isn't going to be very practical.
Most countries have bedrock closer to the surface than that, even if not everywhere. Moreover, mining to more than a mile down isn't too hard, especially if you're not digging through a coal seam (when you would have gas problems). The main issue with deep mines is usually just water ingress, but not all sites have that problem. For example, Boulby (a salt mine) is nearly a mile deep. The only reason we don't normally go down that far is because it's expensive and what we're after is typically closer to the surface.
Except that marijuana smoke does not contain any Tobacco-specific nitrosamines, and the evidence so far is that marijuana smokers are not more likely to develop cancer than non-smokers. That aside, smoking is not the only means by which marijuana is consumed, and non-smoking methods of use appear to have no permanent effects (as opposed to non-smoking methods of using tobacco, which still increase the risk of cancer).
Look, it's hard to believe that putting smoke of any kind in your lungs is actively good for you, whether it comes from tobacco, or marijuana, or even goddamn soft furnishings. As such, if marijuana is legalized then it would be reasonable to regulate its consumption by smoking in the same way as consumption of tobacco through the same route is regulated. Moreover, it has some major effects that are similar to alcohol (slows reaction times, impacts judgement) so there would also be good reason to use some of the same sorts of regulations that are used with alcohol (e.g., don't drive when intoxicated, however you got into that state). These sorts of things cover the key public harms that need to be tackled upon legalization, and the other issues are either none of the governments business anyway (people have a right to harm their own health, even if it is stupid for them to do so) or are a direct consequence of prohibition (the enormous majority of criminality associated with substance use).
In short: legalize and regulate through adaptation of existing rules for other (legal) drugs. You Know It Makes Sense.
Average for an area would be good.
The best would be the mode, i.e., the speed that customers are most likely to see. The median would also be good. They're also not that easy to scam (except by outright lying, which it's a bit too easy to be caught in).
and I am yours, so when you get that check, don't forget to cut me one for $10
You can't get a check. You're an Anonymous Coward.
The median household income in the US is $44k. The mean household income in the US is $60k. Where did you get $31k?
It could have been the modal income; the mode is usually the lowest of the three major average types, for the kinds of distribution seen in things like incomes (where there's a long upper tail but much more bunching up at the bottom end).
Also, the GGP is talking about personal income whereas the GP is talking about household income. The majority of households will have more than one income-earner so things are still consistent.
The median household income in the US is $44k. The mean household income in the US is $60k. Where did you get $31k?
It could have been the modal income; the mode is usually the lowest of the three major average types, for the kinds of distribution seen in things like incomes (where there's a long upper tail but much more bunching up at the bottom end).
Speaking from a statistical PoV. I've no idea what the values actually are. :-)
That's the most insanely impractical philosophy I can imagine. I have to assume you don't actually make anything people use.
Yes, he must be a computer security expert.
...somehow I don't think that a multi-agency-operation can be executed within 1 day
If you show your ass to authorities on six different continents, it goes without saying they're going to feel a lot more generous about cooperating in capturing you.
But it will still take more than a day for them to agree to arrest you. Don't underestimate the need for the management to have a good long meeting first, and the more high-profile agencies involved, the more that becomes necessary. (It's a primate thing, rather like baboons showing off the colour of their genitals.)
I don't think that law would actually hold up in court against plausible deniability encrypted disk images. I think the judge would want proof that a hidden volume actually did exist.
I'd hate to try to use such a defence in a trial. It'd be far more likely that the jury (or triumvirate of magistrates, in a lower-level court) would decide that you were pissing around with them and state that you were guilty, especially if the police provided evidence that you were communicating in a way that indicates that you were involved in hacking. Remember, conviction requires that it be proved beyond reasonable doubt, not beyond all doubt.
This is the UK. Should he be someone from lulzsec and if they have a decent amount of evidence to prove he was a main player I'd say he'll get 2-3 years max and likely out in 12-18 months for good behaviour
At that sentence length, no. Maximum 1/3 off for good behaviour once the sentence is over 2 years long.
You don't need SELinux nor Vsys, you can do that using POSIX file permission. I have a 'firefox' user which can only read write to a 'Downloads' directory, and nothing else.
Now try scaling that up to having two real users on the same system. Or 5 users. Or 20.
Which is why the UK is known for creating all these cloud based services? I must have missed the memo, but nearly all the ones I can think of are from the US. Admittedly, I hear about more of the ones from here because of sample biasing, but I doubt very much that it's that far from the truth.
That's actually mostly sample biasing. UK providers tend to be focused towards selling into either the UK or EU markets, especially dealing with cases where going to the US is utterly not an option, e.g., you'll see very few EU-based firms locating data that contains anything even vaguely personally-identifiable in the US, and that's entirely driven by differences in law, and I've worked with firms that were putting things in the cloud which were even more restricted than that, but again, going to the US wasn't an option, not without proving that Official Secrets were not being exported (we could have proved it for sure, but it was several years of hassle we didn't want).
The UK is also not a particularly attractive place to site a major datacenter: power isn't particularly cheap, and land is (still) very expensive. This encourages UK firms to outsource low-level cloud provision where possible...
Every piece of automation that replaces a person puts that person out of work at that job.
The real question is whether it is better for that person to not be doing that job in the first place. Perhaps they'd be better off overall doing something else. If that's so, then that old job is better automated. (OTOH, just doing them out of a job and leaving them on the dole is not better overall; joblessness is the source of all sorts of social ills. Make-work is better than no-work, even if nowhere near as good as real work.)
Those that do not understand how Public Key Crypto works should not use it.
In other news, gweihir has announced that he will no longer be accessing any website via HTTPS.
(The number of people who understand the whole of a public key crypto system and deployment is vanishingly small. The underlying math is difficult. The programming is easy to make errors in. The way to use it, not all that obvious either going by the massive quantities of misinformation I see here and elsewhere on the 'net. Public key crypto is only practical to use if you don't understand it all; fortunately, there are useful abstractions for most of it that are accessible. Now, if only the firmware makers grokked even that little bit...)
British Rail (which is one of the worst networks in Europe) runs goods trains FASTER than most passenger trains.
It depends on the metric. UK railways tend to run with thinner margins than many other European railways, which means that they've got more opportunity for things to go wrong. The problem is just that UK traffic levels are very high for the amount of number of tracks on any particular route, so failures are much more likely to cause cascading problems. In particular, a lot of routes only have one or two tracks and management hates the thought of investing to fix this.
Look at what has worked and why, look at what hasn't worked and why, but don't look at what hasn't worked and assume that this is a universal constant.
Now that is good advice!
i mean, clearly, emacs is a threat to national security.
Trouble is, they'll ban vi at the same time. Would you want to have to write code in Microsoft Word??? (Shudder)
That's true of reference counting (shared_ptr), not unique_ptr/scoped_ptr, std::array, vector or similar which compile to equivalent C code.
But then you've got objects of the lifespan implied by those entities. If you want to have objects that don't have a tight owning scope or container, then you've got to either put the free/delete in the right place (hard!) or use reference counting, or use garbage collection. You can wrap up lots of stuff around that, but ultimately that's the type of technique you need. That in turn means you've got to pay the cost of the technique. C++'s advantage is that it gives you the flexibility to use a variety of techniques as appropriate. C++'s disadvantage is that you have to select the right technique and it's not guaranteed to be nice if you get it wrong. (Code that runs in a managed environment is more expensive, but much less destructive when things go awry. It's a trade-off.)
Suicide might be technically illegal
Except it is legal in the UK, and has been for decades. Assisting is not legal, mainly to prevent the unscrupulous from manipulating other people into killing themselves. (Some people are utter scum, alas.) It's been stated clearly that while each case of alleged assisting a suicide will be taken on its merits, when it's clear that it's a genuine choice of the suicidee that's not been forced upon them in some way and where there hasn't been too much assistance, there won't be prosecution. That's a wooly position, but just about anything else makes it too easy for the callous (especially if they're family members) to abuse in some way.
Suicide is a terrible thing. It's so unfortunate that it is sometimes the best option. As a society, we should try to ensure that there are as many positive, constructive alternatives to killing oneself as possible. (But if you do decide to go, don't do it by jumping in front of a train or other vehicle; that's very inconsiderate!)