You have a written policy against that kind of thing. You tell employees to remove suchlike should you ever become aware of it, and the responsibility lies with whomever actually did the illegal thing. What a concept !
You're inventing problems that simply don't exist. It's not as if there's any technical barrier to a employee speeding in a company car, calling in bomb-threats from company-phones, hitting someone over the head with a company-owned chair etc etc etc.
Yet in all these cases, the company as such has precisely -zero- responsibility aslong as they did not encourage or assist the crime, or at the very least could be shown to have a policy of silently accepting. (it would, for example, perhaps be different in the case of the speeding if one could show that the company had encouraged employees to speed in order to manage more in a day)
Individual perception counts, but it doesn't dominate.
If you let 100 random men rate 100 random women, you will indeed get several different #1 candidates, and several different #100 candidates.
If you plot the position of one woman though, it will not look anything like random, rather it will have a distinct peak indicating that MOST of the males are close to agreeing MOST of the time.
You won't find a woman that 10 men rate as #1 and 10 other men as #100, instead you'll find that the one rated #1 by most people is generally rated in the top-20 even by those who didn't rate her #1.
There'll be a few ratings way outside the norm, a very few men will rate a woman that most find attractive as unattractive, and vice versa. But those will be few and far between.
The clever thing is that we, the public, can do that too. We always where, offcourse, because the final arbitrator of what was the right and/or wrong choice for a record label or radio-station always was if people where buying the records, or listening to the station.
So, services like last.fm and pandora allow you to NOT wade trough gargantuan heaps of crap to find the stuff you like. By a mechanism not very different from "Customers who bougth this product also bougth..." which should be well-known by now.
The theory is, if you really like music-piece X, Y and Z, and 95% of those that like those three and have also rated W, like W, then odds are very good that you will also like W.
Certainly the odds are MUCH better than "one-size-fits-all" which is what RIAA and major radio-stations deliver. They deliver music that the AVERAGE quite like, and are good at finding music which the maximum amount of people like. But that is a quite different task from finding music that YOU are likely to like.
Total cost to you ? Hit a button now-and-then to indicate "Really dig this track", or "hate this".
You can adjust your sense of adventure too. For example, I currently listen to 50% songs the computer "know" I like (because I've said so) 30% songs the computer thinks I'll very likely like, because the same people tend to like these and others that I like, 10% songs which the computer has no reason to think I like or dislike, but which are in general highly rated, 5% completely random songs and 5% songs which I've said before that I don't like, but that was a long time ago (more than 6 months)
Depends on who you're comparing to offcourse, like always.
I'm sure you're well ahead of the world-average, probably also ahead of the europe-average (though I'm less sure of that truth be told)
But you're miles behind world-leading countries.
Want a piece of my (perfectly average for my geographical area) communications-infrastructure ?
Single-mode-fibre to the basement. Currently modulated at 1Gbps, but physical fibre is capable of another order of magnitude or two, but there's no demand for that. TV over ip, reserved bandwith on the fibre-link so no interference with other stuff. Phone over ip. 25Mbps symetrical internet over fibre. (i.e. 25Mbps upload, 25Mbps download)
That's actually the lowest available speed-tier. (Choices are 25, 50 or 100Mbit/s) Most go for the slowest because frankly, on todays internet "slow" is a relative term and 25Mbit/s is quite acceptable for most uses.
Yeah, it does introduce a single point of failure. There's municipal wi-fi as backup for the internet and gsm-phone as backup for the phones though. No backup for TV, but let's face it, it's seldom life-threathening if TV is down for a few hours. (not that it has ever been down that I can remember)
This ain't crazy good, rather it's a typical hookup for a city. Rural districts mostly don't have fibre and make do with speeds in the 3-20MBit/s range. They also typically do not receive TV over ip.
So, what you're saying is that it'll completely blow away all my files, all my configurations, all my settings, stuff that would take hundreds of hours recreating, if it where possible at all.
But that's not so bad, because the OS, which I can reinstall in an hour, is saved ?
Seriously. The value is ALL in the data, nobody cares if they lose their copy of Ubuntu, it's not as if it's difficult or expensive to get a new copy afterall.
Yeah yeah, backups help. True, but that's unrelated to the STUPID claim that rm -rf ~ is really all that much more benign than rm -rf / would be.
You can have it both ways, or atleast strike a reasonable balance.
For example, you can say it's private, but if the government prevents credible grounds for suspsicion for a judge, the judge can give them a warrant, which again grants access to certain objects or information that they would otherwise not get.
This is what happens to papers in your desk, info from your local bank etc. It's reasonable.
It is however -not- reasonable that swiss banks are exempt from this. There's no particular reason that a warrant against a swiss bank shouldn't be just as effective as one against a non-swiss bank. Except for swiss law, that is.
And that's precisely where the pressure is: A lot of countries put pressure on switzerland to change a set of laws. If they do or not is up to them offcourse.
It's much too small. At the energies involved a micro-black-hole would be aproximately 10^15 times smaller than even a single atom, it'd evaporate instantly.
Even if it did not, it could travel trough solid lead for years before getting lucky and trapping even a single other particle.
So yeah, if it didn't evaporate (which it would) it'd eventually get large enough to cause problems. But more than likely it'd have enough speed to leave the solar-system long before that. And if not, it'd become a problem long after the sun burns out.
Noone argues that hard drugs aren't detrimental to health and happiness. They obviously are. So the first choice would be to have noone addicted to them. But you don't -get- that choice.
Some people ARE addicted to them, that's a fact, and not something it's likely we'll be able to change in the next few years. Given this state of things, the question is how we can minimize the damages caused, to the addicts themselves and society at large.
The reason these should be decriminalized is not that they are beneficial. Overall they clearly are not.
The reason is that the alternative is WORSE.
It is quite silly to essentially say, as society today is; these drugs are bad. Therefore we will let each drug-addict cause a million dollars of damage to society every year, most of it ending in the pockets of criminal gangs, rather than supplying the same addict with the drug at a cost to society of perhaps $2000/year.
The destruction to the addict himself is similar in both cases, the second does have the advantage of not forcing him/her into a life of crime, but the physical and psychiological damages of the drug itself remains.
I could see one argument: If criminalizing the drugs actually led to a significantly lower count of addicts, this would be a great benefit. I don't think it does. Most people would certainly not start using heroin even if they could get it in the drug-store. We've tried this for several decades, and it plain does not seem to be working, even a little bit.
It does however bring literally billions of dollars into the coffers of organized crime.
Making something a drug makes it more expensive, true. But legalizing it makes it much cheaper. Pot is kinda a special case since it's a common plant that grows by itself more or less in most climates. There's a reason it's also known as "weed".
But heroin, cocain and the like cost orders of magnitude over production-cost. Because they are illegal and need to be smuggled in or produced in secret at significant risk.
There are two sides to this, damage to the addicts, and to society. The damage to the addicts is similar if they take similar doses of the same drug, actually probably sligthly lower if legalized because of less overdoses from unknown-strength drugs etc.
Damage to society is today tremendous.
Street-price is somewhere around $100/g, yeah it can vary WILDLY over the map as supply and demand fluctuates, but it's a guesstimate as good as any.
A junkie may consume 2g/day, which works out to $6000/month or thelike, which he/she won't be able to finance legally unless they're well-off, especially since using drugs ain't precisely likely to boost your earnings-potential.
So, there are various low-level crimes commited, by the boatload. Damages are typically MUCH higher than the $6000/month, because replacement-cost is much higher than second-hand value on the black market.
A junkie breaks into your car, damaging the lock in the process, and steals your GPS-unit and stereo. You pay $300 for a new similar GPS, $200 for a new similar stereo and $100 to have the car-lock replaced. A loss of $600, plus the time and annoyance-factor. The junike sells the equipment to some shady character for $75, if that. Having caused 8 times the damage, comapred to the cash gained.
If he/she keeps doing that, the damages caused over a month, just to finance the $6000/month drug-addiction adds up to aproximately $50000/month or $600000/year
That is the cost of a SINGLE junkie that finances the drugs with petty theft. A gargantuan sum.
There's no reason to think heroin should be very different in cost from morphine, if both where legalized. A single user-dose costs something like $0.75 so we're talking $1200/year versus $600000/year, a rather significant difference.
There is a "recommended max speed" of 130 km/h (80.7 mph), this has been the same for atleast the last 15 years or so.
Your insurance will NOT automatically not cover damages even if you go faster, but they will, same as everywhere, try to weasel out of having to pay full damages if the cause of the accident is reckless driving on your part. And offcourse there's a real risk they'll claim that driving faster than the recommended maximum is a reckless thing to do.
If a judge sees it that way or not depends. Driving 120 mph while weaving trough heavy traffic clearly IS reckless. Driving 100 mph under ideal driving-conditions with a good vehicle and low traffic probably would not be.
In practice, most people go 90mph or slower, despite the lack of speed-limits. Most germany still cherish the freedom even if it's a symbolic thing when you yourself are satisfied with 90mph. (which is also hella-fast if you ask me)
There's also the cost-aspect. Driving 120mph rather than 80mph on the parts where that is possible to do safely and legally probably saves you no more than 1/5th the time or something (since it's not unlimited EVERYWHERE, and even where it is you can't ALWAYS safely go 120mph), but it'll cost you double fuel-costs, 80 is already a lot less economical than 60, and the air-friction goes up as the cube of the speed or something similar, you do the math.
I meant the removal of tires. Nobody actually does it, despite the fact that it WOULD decrease the probability that your car is stolen.
Which illustrates things well: Security REALLY ain't worth it when the cost (in money, time, inconvenience) is higher than the expected savings (probability of something happening WITHOUT this security minus probability that it'll happen WITH the security TIMES the cost IF it happens)
The trick is not to eliminate risk, it can't be done. The trick is to search for a strategy that strikes the right balance. It's probably worth it to lock the doors of your car. It may be worth it to have an alarm on it, or to park it in a locked garage. It is CERTAINLY not worth it to remove the tires every time you leave the car for 5 minutes or longer.
But the basic point is healthy. It is perfectly possible to spend so much resources on "security", or to inconvenience people so much trough it that the loss is GREATER than the likely loss from not having that particular piece of "security".
It's not just possible, it's common.
Witness a billion plane-passengers every year stand in security-check lines that are on the average 15 minute longer than they where pre-new-"security". That's 15 billion minutes wasted, plus the cost of the actual security.
That is 2500 lifes wasted yearly, waiting in line. Plus the cost and inconvenience. (what does 15 billion minutes times the average hourly salary of a plane-passenger cost?) How does that compare with the likely added loss of life if security was still at pre-panic levels ? I'm pretty convinced the net-effect is a LOSS.
You're COMPLETELY beside the point. First, my point wasn't "we can't die out cos we're so many", my point was that a HUGE disaster that kills millions or even a billion or two is nevertheless not anywhere close to making us extinct.
Second, the pigeons did not actually DO anything to avoid extinction, or even, for that matter, DISCOVER the threat. They're dumb. Like all animals. We're pretty dumb too, but we're still light-years ahead of any animal on the planet.
Third, I said I don't think its very likely that we'll die out in the next few hundred years, you respond that it's "quite possible", which would've been relevant if I had claimed it was IMPOSSIBLE, which I never did, I just said I don't find it likely.
That humanity will change is a given, such predictions score no points. Indeed even to human beings from a few hundred years back we're half-gods (or atleast we're capable of many feats that THEY assosiate only with Gods), and change is definitely accelerating. (1908 - 2008 is a LOT bigger change than 1208 - 1308 was, really)
That's a given, but it's not precisely extinction, now is it ?
But -some- global disasters may be better dealt with by smarter people.
I don't think it's very likely that humanity will die out in the next few hundred years. It's much more likely that we suffer huge catastrophies of some kind, and that intelligence and resourcefulness determines if 1 million people die or 1 billion.
Which makes a difference, but ain't precisely about extinction seeing as we're several billion people.
It's truly strange though, how "children" are treated in the US. All too frequently as some sort of museum-item, to be protected from the real world and live in a soft, padded, bubble. Frequently literally until they're 18, and then "pop" the bubble is supposed to burst, and the child is supposed to be fully prepared for all the things it was completely shielded from up until the day before.
Okay, so that's caricaturish, but you get the idea.
Why, for example, is it assumed that swearing is bad for children ? INAPROPRIATE swearing is just bad manners, for EVERYONE, regardless of if you're 5 or 50. Bad manners is a problem. Learning to behave is part of growing up, and something the parents should take care of.
There are however situations where swearing is perfectly normal, again, regardless of if you're 5 or 50. I -do- mind if my son would, for example, critisize something his younger sister had made and was proud of as being "shit" (for example). I would however not mind if he said the same thing after say having accidentally dropped a glass that breaks into a million pieces. (I realize norms for behaviour are different in different societies, learning your local norms is however ALSO part of growing up)
As for TV, honestly, if you let the TV raise your children you have problems. Regardless of what whatever censoring-board does or not. How about people take some responsibility for themselves ?
That's all well and good. But in many cases that's a cure worse than the disease. It depends offcourse, on the needs of your workers, the cluefulness of your workers and so on.
Reducing risk is fine, but not if the COST in terms of employee-productivity is a lot higher than any potential win.
Sort of. But on the other hand the companies generally DO settle because the first lawyer they talk to universally tells them the same thing: That they are CLEARLY liable, and that they will DEFINITELY lose in court, so settling is their best bet.
It's not really even a question: BusyBox (and frequently other components) is protected by copyright. To copy and distribute copyrigthed works, one needs a permission from the owner of the copyright, a license. They ain't got one. So they're liable. Plain and simple. Even if the GPL was unenforcable, (which I see no signs of) they'd *STILL* be liable, because without it, they're still lacking any permission to copy and distribute.
It's sort of how, if you aim a gun at someone and ask them to please leave your house, they will generally do so. You may argue it's a pity that this prevents you from establishing the effectiveness of the gun. But really, it works BECAUSE people are generally pretty convinced that being shot with a gun does, infact, hurt.
Plus, it saves the messy cleanup. Shooting people is messy. Court-cases are messy, and they tend to alienate much more than a settlement does.
Personally I think settlements in these cases is win-win. Just like having the person leave WITHOUT you actually needing to pull the trigger is also the best outcome for all involved.
However, since even applications that come included with the OS share the error, the end-result for the user is the same.
Besides, this is a lot like cooperative multitasking.
Saying that deleting open files is possible IF ALL applications cooperate and do the ideal thing is well-and-good, but I'd argue that if you are administrator, then deleting a file should be possible even if the applications are HORRIBLY written.
If nothing else, Windows could say: "File can't be deleted because open by FOO [Terminate foo and delete file] [Cancel] [Mark file for deletion on close]"
But it doesn't. It offers nothing helpful whatsoever. I realize it's possible for people to mess things up by uncritically terminating random shit. I guess this all is a side-effect of the fact that Windows was for a long time such that everyone was logged in as an administrator all the time, so even technical messages meant for admins would in practice be seen by a lot of aunt Tillies.
Furthermore, atleast the major apps delivered by Microsoft themselves should do things correctly, no ? If not it hardly seems reasonable to expect that EVERYONE else should get it right. And infact Office is one of the main culprits in holding open files in modes that prevent them from being deleted.
So use a better tool. The thing about the MS bundleware is that they are all mediocre-at-best additions to the OS. Use something other than Task Manager to kill the process, it'll drop instantly (does for me). There are several quality products that should be on your server. All the Sysinternals come to mind . ..
Needing to install third-party tools for doing trivial routine-tasks like killing a process or deleting a file strikes me as less than optimal. Particularily for people who DONT WANT TO FIDDLE.
I personally don't mind fiddling, and don't care much about windows-annoyances since I use Windows only in rare cases. My point was just to point out that in a number of areas Windows does NOT infact "just work", to the contrary, there are cases where making windows work means fiddling whereas unix "just works". Killing a process and deleting a file are two such examples. Neither are special or rare actions...
Learn your GPOs. You have to do similar type of configuration if you set up Linux outside of Ubuntu (and even then) plus its all graphical. Keep that dialog from showing. You can do pretty much anything with Active Directory,
Again: this answer is completely besides the point when the complaint was that it should JUST WORK with no FIDDLING required. True true, many of the things that DONT work in windows can be made to work, if you're willing to fiddle.
I just love how almost all the complaints with Windows can be fixed with properly configuring the machine. For years I've heard Linux guys say "well if you set it up like this" when the same thing was true with Windows, but they didn't take the time to do it with Windows.
It's a matter of perspective. For years I've been hearing: "With linux you need to fiddle and configure to reach a functional state, with Windows it 'just works'". I'm just pointing out that thats not universally true.
There can be no doubt what the user wishes if he logs in as root/administrator and says: "Kill that process" or "remove that file". In both cases unix "just works" -- it does what the user requests. Windows does NOT. Yes, windows can, with fiddling, be CONVINCED to do as you say. But the out-of-box behaviour is horrible.
Obviously MicroSuck fails for missing the point on their own product goal (eg make it easy for everyone) by doing it poorly, but if your talking computer geek to computer geek the disparity between Windoze and *Nix isn't AS bad as people make it out to be.
This is true. For a nerd, Windows is -somewhat- more annoying, and -somewhat- more of a hassle to get rid of the annoyances on, but most of the truly horrible problems CAN be fiddled-away. The one I haven't found a good solution for is the: Remove the file/directory when I tell you to damnit! problem. If you know a solution, speak.
Yeah, but my point, and the problem in this case is, that as far as I know a file in Windows cannot have -zero- names. On unix that is perfectly possible, though the file is physically removed if the reference-count falls to zero reference counts aren't only directory-entries, but also file-handles.
So, a file in a directory that is open by a process has two references to it. Removing one of them (the directory-entry) poses no problems and can be done at any time. There is still one reference to the file, the file-handle. The process using that file-handle never notices the difference and can carry on reading/writing to it for as long as it care to.
When it finally -does- close the file though, reference-count drops to zero and the file is gone. You don' have to -DO- anything for this to work properly.
But all of which is irrelevant. I'm talking default behaviour here.
On unix, if root says: "rm -f foo.txt", the end-result is that the directory-entry for "foo.txt" is removed from the current directory. This may or may not remove the physical file, depending on if this was the last reference to the file or not.
On Windows, if you select the file and hit "delete" the default behaviour is to say "You can't do that because the file is open" or something to that effect. No option is offered for deleting it on close, informing you WHICH process has it open or "delete it anyway, screw the process".
Yes, it is possible to figure these things out. My point is, it doesn't "just work".
Remember, the entire thing I replied to was a complaint that "I don't want to fiddle, I want it to just work."
When I tell the OS to remove a file, I don't wan't to fiddle. I want it to just work.
Which it does in unix, and does NOT in windows.
That it can be -made- to work by fiddling is sorta beside the point.
Slashdot Mirror
So ?
You have a written policy against that kind of thing. You tell employees to remove suchlike should you ever become aware of it, and the responsibility lies with whomever actually did the illegal thing. What a concept !
You're inventing problems that simply don't exist. It's not as if there's any technical barrier to a employee speeding in a company car, calling in bomb-threats from company-phones, hitting someone over the head with a company-owned chair etc etc etc.
Yet in all these cases, the company as such has precisely -zero- responsibility aslong as they did not encourage or assist the crime, or at the very least could be shown to have a policy of silently accepting. (it would, for example, perhaps be different in the case of the speeding if one could show that the company had encouraged employees to speed in order to manage more in a day)
Sure. But negative is overdoing it; that makes icons and pictures look ridicolous.
Full negative I don't really want. Only a general theme of light on dark background rather than dark on light background
Sure. Donate 1000 copies of some software that is sold retail at $100 but which have marginal cost of production $1.
On paper, you've donated $100,000 worth of software and migth get a $30,000 tax-deduction assuming you pay 30% taxes.
In reality, you've donated something that cost you $1000 to produce, and scored a $30,000 tax-deduction.
I noticed, and I dislike it.
Also, did you notice the trend where it gets harder and harder to run a negative desktop ?
I belong to those that think dark-text-on-brigth-backgrounds is for paper whereas I find the reverse much more readable on a computer-monitor.
Even high-clue applications for technical users, like Eclipse, make it a chore to configure them that way, and I don't get why.
The single exception is CAD-programs. For some reason those all come with easy dark themes, in many of them it's even the default.
Individual perception counts, but it doesn't dominate.
If you let 100 random men rate 100 random women, you will indeed get several different #1 candidates, and several different #100 candidates.
If you plot the position of one woman though, it will not look anything like random, rather it will have a distinct peak indicating that MOST of the males are close to agreeing MOST of the time.
You won't find a woman that 10 men rate as #1 and 10 other men as #100, instead you'll find that the one rated #1 by most people is generally rated in the top-20 even by those who didn't rate her #1.
There'll be a few ratings way outside the norm, a very few men will rate a woman that most find attractive as unattractive, and vice versa. But those will be few and far between.
So ?
..." which should be well-known by now.
The clever thing is that we, the public, can do that too. We always where, offcourse, because the final arbitrator of what was the right and/or wrong choice for a record label or radio-station always was if people where buying the records, or listening to the station.
So, services like last.fm and pandora allow you to NOT wade trough gargantuan heaps of crap to find the stuff you like. By a mechanism not very different from "Customers who bougth this product also bougth
The theory is, if you really like music-piece X, Y and Z, and 95% of those that like those three and have also rated W, like W, then odds are very good that you will also like W.
Certainly the odds are MUCH better than "one-size-fits-all" which is what RIAA and major radio-stations deliver. They deliver music that the AVERAGE quite like, and are good at finding music which the maximum amount of people like. But that is a quite different task from finding music that YOU are likely to like.
Total cost to you ? Hit a button now-and-then to indicate "Really dig this track", or "hate this".
You can adjust your sense of adventure too. For example, I currently listen to 50% songs the computer "know" I like (because I've said so) 30% songs the computer thinks I'll very likely like, because the same people tend to like these and others that I like, 10% songs which the computer has no reason to think I like or dislike, but which are in general highly rated, 5% completely random songs and 5% songs which I've said before that I don't like, but that was a long time ago (more than 6 months)
Try doing -that- with one-size-fits-all.
Depends on who you're comparing to offcourse, like always.
I'm sure you're well ahead of the world-average, probably also ahead of the europe-average (though I'm less sure of that truth be told)
But you're miles behind world-leading countries.
Want a piece of my (perfectly average for my geographical area) communications-infrastructure ?
Single-mode-fibre to the basement. Currently modulated at 1Gbps, but physical fibre is capable of another order of magnitude or two, but there's no demand for that. TV over ip, reserved bandwith on the fibre-link so no interference with other stuff. Phone over ip. 25Mbps symetrical internet over fibre. (i.e. 25Mbps upload, 25Mbps download)
That's actually the lowest available speed-tier. (Choices are 25, 50 or 100Mbit/s) Most go for the slowest because frankly, on todays internet "slow" is a relative term and 25Mbit/s is quite acceptable for most uses.
Yeah, it does introduce a single point of failure. There's municipal wi-fi as backup for the internet and gsm-phone as backup for the phones though. No backup for TV, but let's face it, it's seldom life-threathening if TV is down for a few hours. (not that it has ever been down that I can remember)
This ain't crazy good, rather it's a typical hookup for a city. Rural districts mostly don't have fibre and make do with speeds in the 3-20MBit/s range. They also typically do not receive TV over ip.
So, what you're saying is that it'll completely blow away all my files, all my configurations, all my settings, stuff that would take hundreds of hours recreating, if it where possible at all.
But that's not so bad, because the OS, which I can reinstall in an hour, is saved ?
Seriously. The value is ALL in the data, nobody cares if they lose their copy of Ubuntu, it's not as if it's difficult or expensive to get a new copy afterall.
Yeah yeah, backups help. True, but that's unrelated to the STUPID claim that rm -rf ~ is really all that much more benign than rm -rf / would be.
Yeah, I agree. Absent working lasting treatment, junkies will still cause damage and do the same harmful things they do today.
If -ALL- you do is give them heroin (more practically metadon) in the drugstore, then ALL you achieve is:
a) Somewhat increased safety for the junkie due to known strength, quality-tested drugs.
b) About half a million dollars damages to society avoided every year pro junkie.
c) About half a million dollars less financing for criminal groups every year, pro junkie.
The person will as you say likely still be a net drain on society, just a SMALLER drain than he/she used to be.
You can have it both ways, or atleast strike a reasonable balance.
For example, you can say it's private, but if the government prevents credible grounds for suspsicion for a judge, the judge can give them a warrant, which again grants access to certain objects or information that they would otherwise not get.
This is what happens to papers in your desk, info from your local bank etc. It's reasonable.
It is however -not- reasonable that swiss banks are exempt from this. There's no particular reason that a warrant against a swiss bank shouldn't be just as effective as one against a non-swiss bank. Except for swiss law, that is.
And that's precisely where the pressure is: A lot of countries put pressure on switzerland to change a set of laws. If they do or not is up to them offcourse.
It's much too small. At the energies involved a micro-black-hole would be aproximately 10^15 times smaller than even a single atom, it'd evaporate instantly.
Even if it did not, it could travel trough solid lead for years before getting lucky and trapping even a single other particle.
So yeah, if it didn't evaporate (which it would) it'd eventually get large enough to cause problems. But more than likely it'd have enough speed to leave the solar-system long before that. And if not, it'd become a problem long after the sun burns out.
Noone argues that hard drugs aren't detrimental to health and happiness. They obviously are. So the first choice would be to have noone addicted to them. But you don't -get- that choice.
Some people ARE addicted to them, that's a fact, and not something it's likely we'll be able to change in the next few years. Given this state of things, the question is how we can minimize the damages caused, to the addicts themselves and society at large.
The reason these should be decriminalized is not that they are beneficial. Overall they clearly are not.
The reason is that the alternative is WORSE.
It is quite silly to essentially say, as society today is; these drugs are bad. Therefore we will let each drug-addict cause a million dollars of damage to society every year, most of it ending in the pockets of criminal gangs, rather than supplying the same addict with the drug at a cost to society of perhaps $2000/year.
The destruction to the addict himself is similar in both cases, the second does have the advantage of not forcing him/her into a life of crime, but the physical and psychiological damages of the drug itself remains.
I could see one argument: If criminalizing the drugs actually led to a significantly lower count of addicts, this would be a great benefit. I don't think it does. Most people would certainly not start using heroin even if they could get it in the drug-store. We've tried this for several decades, and it plain does not seem to be working, even a little bit.
It does however bring literally billions of dollars into the coffers of organized crime.
Making something a drug makes it more expensive, true. But legalizing it makes it much cheaper. Pot is kinda a special case since it's a common plant that grows by itself more or less in most climates. There's a reason it's also known as "weed".
But heroin, cocain and the like cost orders of magnitude over production-cost. Because they are illegal and need to be smuggled in or produced in secret at significant risk.
There are two sides to this, damage to the addicts, and to society. The damage to the addicts is similar if they take similar doses of the same drug, actually probably sligthly lower if legalized because of less overdoses from unknown-strength drugs etc.
Damage to society is today tremendous.
Street-price is somewhere around $100/g, yeah it can vary WILDLY over the map as supply and demand fluctuates, but it's a guesstimate as good as any.
A junkie may consume 2g/day, which works out to $6000/month or thelike, which he/she won't be able to finance legally unless they're well-off, especially since using drugs ain't precisely likely to boost your earnings-potential.
So, there are various low-level crimes commited, by the boatload. Damages are typically MUCH higher than the $6000/month, because replacement-cost is much higher than second-hand value on the black market.
A junkie breaks into your car, damaging the lock in the process, and steals your GPS-unit and stereo. You pay $300 for a new similar GPS, $200 for a new similar stereo and $100 to have the car-lock replaced. A loss of $600, plus the time and annoyance-factor. The junike sells the equipment to some shady character for $75, if that. Having caused 8 times the damage, comapred to the cash gained.
If he/she keeps doing that, the damages caused over a month, just to finance the $6000/month drug-addiction adds up to aproximately $50000/month or $600000/year
That is the cost of a SINGLE junkie that finances the drugs with petty theft. A gargantuan sum.
There's no reason to think heroin should be very different in cost from morphine, if both where legalized. A single user-dose costs something like $0.75 so we're talking $1200/year versus $600000/year, a rather significant difference.
Actually, you're misremembering.
There is a "recommended max speed" of 130 km/h (80.7 mph), this has been the same for atleast the last 15 years or so.
Your insurance will NOT automatically not cover damages even if you go faster, but they will, same as everywhere, try to weasel out of having to pay full damages if the cause of the accident is reckless driving on your part. And offcourse there's a real risk they'll claim that driving faster than the recommended maximum is a reckless thing to do.
If a judge sees it that way or not depends. Driving 120 mph while weaving trough heavy traffic clearly IS reckless. Driving 100 mph under ideal driving-conditions with a good vehicle and low traffic probably would not be.
In practice, most people go 90mph or slower, despite the lack of speed-limits. Most germany still cherish the freedom even if it's a symbolic thing when you yourself are satisfied with 90mph. (which is also hella-fast if you ask me)
There's also the cost-aspect. Driving 120mph rather than 80mph on the parts where that is possible to do safely and legally probably saves you no more than 1/5th the time or something (since it's not unlimited EVERYWHERE, and even where it is you can't ALWAYS safely go 120mph), but it'll cost you double fuel-costs, 80 is already a lot less economical than 60, and the air-friction goes up as the cube of the speed or something similar, you do the math.
I meant the removal of tires. Nobody actually does it, despite the fact that it WOULD decrease the probability that your car is stolen.
Which illustrates things well: Security REALLY ain't worth it when the cost (in money, time, inconvenience) is higher than the expected savings (probability of something happening WITHOUT this security minus probability that it'll happen WITH the security TIMES the cost IF it happens)
The trick is not to eliminate risk, it can't be done. The trick is to search for a strategy that strikes the right balance. It's probably worth it to lock the doors of your car. It may be worth it to have an alarm on it, or to park it in a locked garage. It is CERTAINLY not worth it to remove the tires every time you leave the car for 5 minutes or longer.
Your example is contrived.
But the basic point is healthy. It is perfectly possible to spend so much resources on "security", or to inconvenience people so much trough it that the loss is GREATER than the likely loss from not having that particular piece of "security".
It's not just possible, it's common.
Witness a billion plane-passengers every year stand in security-check lines that are on the average 15 minute longer than they where pre-new-"security". That's 15 billion minutes wasted, plus the cost of the actual security.
That is 2500 lifes wasted yearly, waiting in line. Plus the cost and inconvenience. (what does 15 billion minutes times the average hourly salary of a plane-passenger cost?) How does that compare with the likely added loss of life if security was still at pre-panic levels ? I'm pretty convinced the net-effect is a LOSS.
You're COMPLETELY beside the point. First, my point wasn't "we can't die out cos we're so many", my point was that a HUGE disaster that kills millions or even a billion or two is nevertheless not anywhere close to making us extinct.
Second, the pigeons did not actually DO anything to avoid extinction, or even, for that matter, DISCOVER the threat. They're dumb. Like all animals. We're pretty dumb too, but we're still light-years ahead of any animal on the planet.
Third, I said I don't think its very likely that we'll die out in the next few hundred years, you respond that it's "quite possible", which would've been relevant if I had claimed it was IMPOSSIBLE, which I never did, I just said I don't find it likely.
That humanity will change is a given, such predictions score no points. Indeed even to human beings from a few hundred years back we're half-gods (or atleast we're capable of many feats that THEY assosiate only with Gods), and change is definitely accelerating. (1908 - 2008 is a LOT bigger change than 1208 - 1308 was, really)
That's a given, but it's not precisely extinction, now is it ?
True enough.
But -some- global disasters may be better dealt with by smarter people.
I don't think it's very likely that humanity will die out in the next few hundred years. It's much more likely that we suffer huge catastrophies of some kind, and that intelligence and resourcefulness determines if 1 million people die or 1 billion.
Which makes a difference, but ain't precisely about extinction seeing as we're several billion people.
It's truly strange though, how "children" are treated in the US. All too frequently as some sort of museum-item, to be protected from the real world and live in a soft, padded, bubble. Frequently literally until they're 18, and then "pop" the bubble is supposed to burst, and the child is supposed to be fully prepared for all the things it was completely shielded from up until the day before.
Okay, so that's caricaturish, but you get the idea.
Why, for example, is it assumed that swearing is bad for children ? INAPROPRIATE swearing is just bad manners, for EVERYONE, regardless of if you're 5 or 50. Bad manners is a problem. Learning to behave is part of growing up, and something the parents should take care of.
There are however situations where swearing is perfectly normal, again, regardless of if you're 5 or 50. I -do- mind if my son would, for example, critisize something his younger sister had made and was proud of as being "shit" (for example). I would however not mind if he said the same thing after say having accidentally dropped a glass that breaks into a million pieces. (I realize norms for behaviour are different in different societies, learning your local norms is however ALSO part of growing up)
As for TV, honestly, if you let the TV raise your children you have problems. Regardless of what whatever censoring-board does or not. How about people take some responsibility for themselves ?
That's all well and good. But in many cases that's a cure worse than the disease. It depends offcourse, on the needs of your workers, the cluefulness of your workers and so on.
Reducing risk is fine, but not if the COST in terms of employee-productivity is a lot higher than any potential win.
Like always, "it depends".
Sort of. But on the other hand the companies generally DO settle because the first lawyer they talk to universally tells them the same thing: That they are CLEARLY liable, and that they will DEFINITELY lose in court, so settling is their best bet.
It's not really even a question: BusyBox (and frequently other components) is protected by copyright. To copy and distribute copyrigthed works, one needs a permission from the owner of the copyright, a license. They ain't got one. So they're liable. Plain and simple. Even if the GPL was unenforcable, (which I see no signs of) they'd *STILL* be liable, because without it, they're still lacking any permission to copy and distribute.
It's sort of how, if you aim a gun at someone and ask them to please leave your house, they will generally do so. You may argue it's a pity that this prevents you from establishing the effectiveness of the gun. But really, it works BECAUSE people are generally pretty convinced that being shot with a gun does, infact, hurt.
Plus, it saves the messy cleanup. Shooting people is messy. Court-cases are messy, and they tend to alienate much more than a settlement does.
Personally I think settlements in these cases is win-win. Just like having the person leave WITHOUT you actually needing to pull the trigger is also the best outcome for all involved.
Perhaps you're technically correct.
However, since even applications that come included with the OS share the error, the end-result for the user is the same.
Besides, this is a lot like cooperative multitasking.
Saying that deleting open files is possible IF ALL applications cooperate and do the ideal thing is well-and-good, but I'd argue that if you are administrator, then deleting a file should be possible even if the applications are HORRIBLY written.
If nothing else, Windows could say: "File can't be deleted because open by FOO [Terminate foo and delete file] [Cancel] [Mark file for deletion on close]"
But it doesn't. It offers nothing helpful whatsoever. I realize it's possible for people to mess things up by uncritically terminating random shit. I guess this all is a side-effect of the fact that Windows was for a long time such that everyone was logged in as an administrator all the time, so even technical messages meant for admins would in practice be seen by a lot of aunt Tillies.
Furthermore, atleast the major apps delivered by Microsoft themselves should do things correctly, no ? If not it hardly seems reasonable to expect that EVERYONE else should get it right. And infact Office is one of the main culprits in holding open files in modes that prevent them from being deleted.
So use a better tool. The thing about the MS bundleware is that they are all mediocre-at-best additions to the OS. Use something other than Task Manager to kill the process, it'll drop instantly (does for me). There are several quality products that should be on your server. All the Sysinternals come to mind . . .
Needing to install third-party tools for doing trivial routine-tasks like killing a process or deleting a file strikes me as less than optimal. Particularily for people who DONT WANT TO FIDDLE.
I personally don't mind fiddling, and don't care much about windows-annoyances since I use Windows only in rare cases. My point was just to point out that in a number of areas Windows does NOT infact "just work", to the contrary, there are cases where making windows work means fiddling whereas unix "just works". Killing a process and deleting a file are two such examples. Neither are special or rare actions...
Learn your GPOs. You have to do similar type of configuration if you set up Linux outside of Ubuntu (and even then) plus its all graphical. Keep that dialog from showing. You can do pretty much anything with Active Directory,
Again: this answer is completely besides the point when the complaint was that it should JUST WORK with no FIDDLING required. True true, many of the things that DONT work in windows can be made to work, if you're willing to fiddle.
I just love how almost all the complaints with Windows can be fixed with properly configuring the machine. For years I've heard Linux guys say "well if you set it up like this" when the same thing was true with Windows, but they didn't take the time to do it with Windows.
It's a matter of perspective. For years I've been hearing: "With linux you need to fiddle and configure to reach a functional state, with Windows it 'just works'". I'm just pointing out that thats not universally true.
There can be no doubt what the user wishes if he logs in as root/administrator and says: "Kill that process" or "remove that file". In both cases unix "just works" -- it does what the user requests. Windows does NOT. Yes, windows can, with fiddling, be CONVINCED to do as you say. But the out-of-box behaviour is horrible.
Obviously MicroSuck fails for missing the point on their own product goal (eg make it easy for everyone) by doing it poorly, but if your talking computer geek to computer geek the disparity between Windoze and *Nix isn't AS bad as people make it out to be.
This is true. For a nerd, Windows is -somewhat- more annoying, and -somewhat- more of a hassle to get rid of the annoyances on, but most of the truly horrible problems CAN be fiddled-away. The one I haven't found a good solution for is the: Remove the file/directory when I tell you to damnit! problem. If you know a solution, speak.
I know. But that's not really a good answer if the problem was: "I don't want to fiddle, I want it to JUST WORK", no ?
True, many windows-annoyances can be removed if you ARE willing to fiddle.
But that's little comfort if your reason for using Windows is that you DONT want to fiddle.
Yeah, but my point, and the problem in this case is, that as far as I know a file in Windows cannot have -zero- names. On unix that is perfectly possible, though the file is physically removed if the reference-count falls to zero reference counts aren't only directory-entries, but also file-handles.
So, a file in a directory that is open by a process has two references to it. Removing one of them (the directory-entry) poses no problems and can be done at any time. There is still one reference to the file, the file-handle. The process using that file-handle never notices the difference and can carry on reading/writing to it for as long as it care to.
When it finally -does- close the file though, reference-count drops to zero and the file is gone. You don' have to -DO- anything for this to work properly.
But all of which is irrelevant. I'm talking default behaviour here.
On unix, if root says: "rm -f foo.txt", the end-result is that the directory-entry for "foo.txt" is removed from the current directory. This may or may not remove the physical file, depending on if this was the last reference to the file or not.
On Windows, if you select the file and hit "delete" the default behaviour is to say "You can't do that because the file is open" or something to that effect. No option is offered for deleting it on close, informing you WHICH process has it open or "delete it anyway, screw the process".
Yes, it is possible to figure these things out. My point is, it doesn't "just work".
Remember, the entire thing I replied to was a complaint that "I don't want to fiddle, I want it to just work."
When I tell the OS to remove a file, I don't wan't to fiddle. I want it to just work.
Which it does in unix, and does NOT in windows.
That it can be -made- to work by fiddling is sorta beside the point.