What you say is what she is now claiming to have meant, after the whole of Italy has been laughing at her. Unfortunately for her, and for us all Italians, there is no possible denying, what was written in the press release literally and unequivocally means "tunnel between CERN and Gran Sasso, across which the experiment took place".
There is no possible misinterpretation, and the phrase literally means "To the construction of the tunnel between CERN and the Gran Sasso laboratories". No cultural context, definition or whatever. "Tunnel" is a commonly used word in Italian, with the exact same meaning - an approximately horizontal hole in the ground.
I second the comments of Nite_Hawk. I should add that the other big question you'll face is monitoring and alerting; look at Ganglia and Nagios/Icinga.
Our team manages the Online/DAQ system of the ATLAS experiment at CERN, about 2000 PCs. Central configuration is of course a must; we use Scientific Linux CERN 5 and netbooted, single image for most "worker" nodes, with specialisation done after boot by an in-house scripting system; servers and other special purpose systems are managed via Quattor or Puppet. Quattor is used a lot at CERN, I find that it is less flexible and requires more work than Puppet but its (currently) stronger package management gives better guarantees of uniformity. Netbooting a single image is excellent to guarantee that all systems are the same and you just need to reboot to be like on a new clean install; but it's a lot of work to setup and maintain, so it's only justified over large numbers.
Personally I also manage the grid cluster at University of Johannesburg (~30 nodes) and there I'm using Cobbler and Puppet, they do a nice job and don't take too much effort.
To make antihydrogen, the accelerators that feed protons to the Large Hadron Collider (LHC) at CERN divert some of these to make antiprotons by slamming them into a metal target
On the same vein, all CERN things could also be said to be from EDF scientists, since CERN is powered by electrons coming from EDF (the French electricity company).
In the case of my bank, the generated key is only required when operating on the account - making a payment, changing personal data etc.
I consider this actually better, because you get used to never have to give all of them together - which would allow a window of opportunity for the keylogger to use your temp password to login to the real account and make transactions from it.
User interface is largely irrelevant. As long as the controls you need are there, you can work it out.
Sure the technical capabilities are the main issue, but some scopes have really awkward interfaces, which slow you down badly even after you have got used to them. And I don't know about you, but I often happen to be in an hurry...:-)
Interestingly, the bad interface is not only an issue with the cheaper, odd-brand scopes, but also with some very high-level ones, which end up being too fancy and with too many configurations and settings. At one of the labs where I work we have a great 2GHz Tektronik with large sample buffers etc, but I only use it when I need to store fast waveforms or take statistical measurements; for day-to-day work the simpler ones are much quicker to use.
My understanding is that the big problem was that the pressure relief valves of the external vacuum vessels had been designed to handle the case of a "normal" helium leak or of a quench, not the huge amount of helium gas plus heating from the spark plus the simultaneous quench of a 100 neighbouring magnets. If the pressure did not build up, the damage would have been far more limited.
So they are going to install additional and/or larger pressure relief valves on each magnet. It's a huge relief that the problem was not in the inherent design, and the fix is a relatively simple modification. Still a very big job though.
Dividing into more sections would have increased the complexity and costs even further, and it would not really have increased the reliability - more piping and cabling, more things which can fail.
In any case, even if each single magnet was separate, it would still take weeks for warm-up and cool-down, so you would not gain much.
Which makes me wonder -- shouldn't they have oxygen masks available around the accelerator in case something like this happens?
Not only there are some of those, but most of all everybody carries a personal breathing apparatus, when they go down in the experiment pits or in the tunnel. You wouldn't want tens of people to fight over a few masks.
If you work in any experimental area at CERN, you must pass some safety course, and if you work in the LHC underground areas there is further specific training do to.
In any case, when the beam is on,which is when the risk of this kind of failure is highest, nobody can stay in the tunnel because of the radiation levels.
No, don't make vacuum. The low pressure will make plastics outgas their solvents/plasticisers faster, and become brittle. An inert gas like nitrogen should be better.
The entire USA civil legal industry is based on the fact that it costs far more to defend yourself against whatever the plaintiff claim (lie) than to pay what they ask. Plaintiffs are almost never forced to pay the legal fees of the defendant, unless the case is very public AND black and white. It is all a sham and a huge subsidy for sleazy attorneys that know how to work the system, often at the expense of an insurance company, but not always. I found this out the hard way when I made the mistake of selling my home to a sleazeball attorney. They can fuck with you based on the most flimsy of reasons and it costs them very little to ruin your life. The defense attorneys, that burn through their client's life savings by over billing and accomplishing nothing, but still make costly mistakes, are no better. "Justice" in the USA is only for the rich. Far worse than the money, is the stress, the fear that my children may not be able to attend college because of it. It should be a crime, but it never will. Who runs the USA court system? Judges, that are also attorneys. Who makes the USA laws? Elected attorneys.
There, fixed for you. Other countries have different problems.
oh and btw, I've yet to see the first magnetic memory in operation. Yeah, me too - the only one I've ever seen was gathering dust on a shelf of the Physics datacentre, 15 years ago. And I haven't seen puch cards in operation either!
It's so nice those few times when technology makes me feel young:-P
I really enjoyed Anonymous Coward trying to be funny, perhaps. Guess ACs aren't as smart as they think after all, next thing they'll be saying that universe is electromagnetic. It's about time these ACs were held accountable, I'm tired of being tethered to their pet theories. They are so holding us back.
The solution is actually quite simple: use cheap incandescent bulbs where the light is often switched on/off, like home corridors (this also have the advantage of immediately giving maximum lighting, and power consumption is almost negligible); use fluorescent ones in those rooms where you spend longer periods of time.
Also, it's better to avoid the very cheapest CFDs, as the electronics inside them is of low quality and the failure rate is somewhat higher.
But spin-down/spin-up is possibly the most stressful procedure for an HD, so you don't want to do it too often. If you are not convinced, just note that one of the "old-age" parameters in SMART is Start_Stop_Count.
Pray, cancel what, if there is no contract?
No, seriously, the whole point about having contracts (and never mind if they need signing or not - see shrink-wrap EULAs) is that both parties agree on rights and duties. Saying that there is no contract, you have actually confirmed my point: the customer is given no rights, it's a Wild West kind of law of the mightier. The picture you are portraying would be completely illegal anywhere in Europe - probably here in South Africa too.
Technically, I understand your point, 60Gb/month does sound like a lot to me too - but why not spell it out in a contract, or Service Agreement, whatever you want to call it? How would that hurt you?
"We have to draw the line somewhere," NO, not SOMEWHERE, but in the CONTRACT.
That's what all the USA people in this thread are angry about. How can you not understand this? All your justifications, no matter how reasonable they may sound, mean nothing. It's just one more bit of evidence that in the USA companies can do whatever they want, and customers have no rights.
I'm any italian, but living in South Africa since almost two years. Being stuck with a 3GB/month cap was almost a shock to me (when there's a new DVD ISO of the Linux I use I can only download it in the weekend, at the university) but at least Telkom SA is CLEAR in their contract about the limits.
Slashdot Mirror
What you say is what she is now claiming to have meant, after the whole of Italy has been laughing at her. Unfortunately for her, and for us all Italians, there is no possible denying, what was written in the press release literally and unequivocally means "tunnel between CERN and Gran Sasso, across which the experiment took place".
There is no possible misinterpretation, and the phrase literally means "To the construction of the tunnel between CERN and the Gran Sasso laboratories". No cultural context, definition or whatever. "Tunnel" is a commonly used word in Italian, with the exact same meaning - an approximately horizontal hole in the ground.
I second the comments of Nite_Hawk. I should add that the other big question you'll face is monitoring and alerting; look at Ganglia and Nagios/Icinga.
Our team manages the Online/DAQ system of the ATLAS experiment at CERN, about 2000 PCs. Central configuration is of course a must; we use Scientific Linux CERN 5 and netbooted, single image for most "worker" nodes, with specialisation done after boot by an in-house scripting system; servers and other special purpose systems are managed via Quattor or Puppet. Quattor is used a lot at CERN, I find that it is less flexible and requires more work than Puppet but its (currently) stronger package management gives better guarantees of uniformity. Netbooting a single image is excellent to guarantee that all systems are the same and you just need to reboot to be like on a new clean install; but it's a lot of work to setup and maintain, so it's only justified over large numbers.
Personally I also manage the grid cluster at University of Johannesburg (~30 nodes) and there I'm using Cobbler and Puppet, they do a nice job and don't take too much effort.
PS shmux is your friend :-)
To make antihydrogen, the accelerators that feed protons to the Large Hadron Collider (LHC) at CERN divert some of these to make antiprotons by slamming them into a metal target
On the same vein, all CERN things could also be said to be from EDF scientists, since CERN is powered by electrons coming from EDF (the French electricity company).
Of course. Tigers are outdated. Leopard's the name of the new threat.
In the case of my bank, the generated key is only required when operating on the account - making a payment, changing personal data etc.
I consider this actually better, because you get used to never have to give all of them together - which would allow a window of opportunity for the keylogger to use your temp password to login to the real account and make transactions from it.
User interface is largely irrelevant. As long as the controls you need are there, you can work it out.
Sure the technical capabilities are the main issue, but some scopes have really awkward interfaces, which slow you down badly even after you have got used to them. And I don't know about you, but I often happen to be in an hurry... :-)
Interestingly, the bad interface is not only an issue with the cheaper, odd-brand scopes, but also with some very high-level ones, which end up being too fancy and with too many configurations and settings. At one of the labs where I work we have a great 2GHz Tektronik with large sample buffers etc, but I only use it when I need to store fast waveforms or take statistical measurements; for day-to-day work the simpler ones are much quicker to use.
My understanding is that the big problem was that the pressure relief valves of the external vacuum vessels had been designed to handle the case of a "normal" helium leak or of a quench, not the huge amount of helium gas plus heating from the spark plus the simultaneous quench of a 100 neighbouring magnets. If the pressure did not build up, the damage would have been far more limited.
So they are going to install additional and/or larger pressure relief valves on each magnet. It's a huge relief that the problem was not in the inherent design, and the fix is a relatively simple modification. Still a very big job though.
~25 years, 0 exploding, 2 MBs (from the same batch) with leaking caps.
Dividing into more sections would have increased the complexity and costs even further, and it would not really have increased the reliability - more piping and cabling, more things which can fail.
In any case, even if each single magnet was separate, it would still take weeks for warm-up and cool-down, so you would not gain much.
Which makes me wonder -- shouldn't they have oxygen masks available around the accelerator in case something like this happens?
Not only there are some of those, but most of all everybody carries a personal breathing apparatus, when they go down in the experiment pits or in the tunnel. You wouldn't want tens of people to fight over a few masks.
If you work in any experimental area at CERN, you must pass some safety course, and if you work in the LHC underground areas there is further specific training do to.
In any case, when the beam is on,which is when the risk of this kind of failure is highest, nobody can stay in the tunnel because of the radiation levels.
No, don't make vacuum. The low pressure will make plastics outgas their solvents/plasticisers faster, and become brittle. An inert gas like nitrogen should be better.
The entire USA civil legal industry is based on the fact that it costs far more to defend yourself against whatever the plaintiff claim (lie) than to pay what they ask. Plaintiffs are almost never forced to pay the legal fees of the defendant, unless the case is very public AND black and white. It is all a sham and a huge subsidy for sleazy attorneys that know how to work the system, often at the expense of an insurance company, but not always. I found this out the hard way when I made the mistake of selling my home to a sleazeball attorney. They can fuck with you based on the most flimsy of reasons and it costs them very little to ruin your life. The defense attorneys, that burn through their client's life savings by over billing and accomplishing nothing, but still make costly mistakes, are no better. "Justice" in the USA is only for the rich. Far worse than the money, is the stress, the fear that my children may not be able to attend college because of it. It should be a crime, but it never will. Who runs the USA court system? Judges, that are also attorneys. Who makes the USA laws? Elected attorneys.
There, fixed for you. Other countries have different problems.
They must suffer personally and professionally for this, not just in the form of business losses.
Where "They" should mean both the RIAA and its lawyers.
Yes, the lawyers, they also should be held responsible for their actions. This kind of nonsense will never stop otherwise.
Isn't that called LGPL already?
Alice, welcome to Dilbertland, come see how deep the rabbit hole is...
Really, who would actually work for a company that hires you after having said something like this?
It's so nice those few times when technology makes me feel young
(I'm tempted to call it crackpot theory but will leave that to an actual physicist)
Yes, it's a crackpot theory.- sash, physicist.
(yes, mazarin5 came first on this, but I couldn't resist anyway...)
I really enjoyed Anonymous Coward trying to be funny, perhaps. Guess ACs aren't as smart as they think after all, next thing they'll be saying that universe is electromagnetic. It's about time these ACs were held accountable, I'm tired of being tethered to their pet theories. They are so holding us back.
The solution is actually quite simple: use cheap incandescent bulbs where the light is often switched on/off, like home corridors (this also have the advantage of immediately giving maximum lighting, and power consumption is almost negligible); use fluorescent ones in those rooms where you spend longer periods of time.
Also, it's better to avoid the very cheapest CFDs, as the electronics inside them is of low quality and the failure rate is somewhat higher.
But spin-down/spin-up is possibly the most stressful procedure for an HD, so you don't want to do it too often. If you are not convinced, just note that one of the "old-age" parameters in SMART is Start_Stop_Count.
Pray, cancel what, if there is no contract?
No, seriously, the whole point about having contracts (and never mind if they need signing or not - see shrink-wrap EULAs) is that both parties agree on rights and duties. Saying that there is no contract, you have actually confirmed my point: the customer is given no rights, it's a Wild West kind of law of the mightier. The picture you are portraying would be completely illegal anywhere in Europe - probably here in South Africa too.
Technically, I understand your point, 60Gb/month does sound like a lot to me too - but why not spell it out in a contract, or Service Agreement, whatever you want to call it? How would that hurt you?
I still prefer to think that we just haven't found yet that solution that sucks less, and we should keep thinking on.
"We have to draw the line somewhere," NO, not SOMEWHERE, but in the CONTRACT.
That's what all the USA people in this thread are angry about. How can you not understand this? All your justifications, no matter how reasonable they may sound, mean nothing. It's just one more bit of evidence that in the USA companies can do whatever they want, and customers have no rights.
I'm any italian, but living in South Africa since almost two years. Being stuck with a 3GB/month cap was almost a shock to me (when there's a new DVD ISO of the Linux I use I can only download it in the weekend, at the university) but at least Telkom SA is CLEAR in their contract about the limits.