Cash is not necessarily cheaper for the merchant. The cost of secure transport of cash to the bank can be the same as merchant fees. On reason some supermarkets offer cashback - it reduces their cash transport costs.
No, that is what Hawking, and a considerable number of other scientists believe. Essentially, nature is allowed to "borrow" energy from nowhere provided the product of the energy and time the energy exists does not exceed Planks constant. When it does so, a particle and its matching antiparticle (to keep all the charges, baryon numbers etc. matched) spring into existence for a very short time, then cancel out again, "repaying" the borrowed energy.
Except that if this happens really close to the event horizon of a black hole, one of the two particles can fall into the hole and the other doesn't, resulting in the net creation of a particle outside the event horizon. The energy needed to "balance the books" and create the particle comes from the black hole. This means that black holes are continuously emitting particles, which are called Hawking Radiation, and losing energy. However, to maximise the chance of one particle falling in and the other escaping, the gravitational field has to be very non-linear, which means that the hole has to be small. The smaller the hole, the faster it evaporates, so the faster it shrinks which eventually leads to a runaway; tiny black holes explode. However, stellar mass black holes evaporate so slowly that it takes a bucket load of exponents to measure the time until they explode.
I have used it for successfully to get out at the right stop when I cannot read the station names. You get out when the train is scheduled to arrive at your destination. My experience is that (a) the train is actually stopped within the scheduled minute, and (b) it is at the right destination. Very comforting when the script it complete gibberish to you.
This is probably a problem in jobs where people want interchangeable droids. But the sort of jobs wanted by the sort of people who read/. want people with originality and flexibility. And if you are original and flexible, you are probably going to have an original and flexible private life. So if an employer is going to get upset about the somewhat odd (but legal) things you may get up to at home, they are probably so strait-laced that you don't want to work for them, and they are probably not going to succeed in the long term. (If you do illegal things AND post them to the Net, you are too stupid to deserve the job.)
Certainly a net search on my name reveals a lot about me - very boring, mostly, but accurate. But if any employer objects to any of the trail I have left, I probably don't want to work for them anyway. Of course, when the economy crumbles to the point I am reduced to burger flipping, I may change my mind.
Also, that is machine code, not SQL. SQL has an built-in date type, whose internal representation is opaque, with which it is capable of doing date arithmetic e.g. finding the number of days between two dates expressed, in human terms, as year/month/day. One of the points of using a proper database it to avoid having to re-invent all the permutations of date arithmetic.
You could perfectly well allocate magic dates for special meanings; this would be regarded by most database professionals as the dirtiest of kludges.
I said not that it means anything in SQL. I agree that it means "absence of data". But a database designer must specify what "absence of data" means in the context of this table, just as a C programmer must know what "not pointing to anything" (aka a C NULL) means in his context. Does "absence of data" in the Date of Death column mean the person is alive, or that the database does not know his death date? A matter of design, not programming.
How much memory will you pay to avoid one bug? Null terminated strings were invented when 64k was a large memory for a single computer. We have moved rather a long way from that. I would spend a *lot* of ram to save one bug in delivered code.
Which is solved if lists and trees are implemented within the system in the same way as integers and floats. Lists are implemented in Python. The article was referring to language design: I think most language designers today would ensure that lists, both fixed and variable, were either in the language or easily implemented in efficient library functions. Trees don't seem to have the same near-universal acceptance.
What NULL "means" is undefined. There are at least two possible meanings, and it is up to the database designer to define, for every column in which null is allowed. The two meanings obvious meanings are either "unknown" or "has not got". For a Date Of Birth column, it is obviously "unknown", because everybody has a DOB. But what does a null Date Of Death column mean? In a commercial context, it probably means that, so far as the database is concerned, the person is still alive. But in a historical database, it could mean that the DOD is really unknown, or that the person is alive. Likewise an Address3 line might interpret NULL as Address3 is "unknown" if Address1 is null and "has not got" if Address1 is non-null. Good database design will always specify this.
There are some places - the Scottish Islands are one - where the crime rate is low enough that people routinely leave their houses unlocked. Neighbours can enter the house, e.g. to borrow and return things, at will. Likewise car keys are left in the ignition so that if the car is in the way anybody can move it.
I realise that it is impossibly idealistic to expect this to work in cities. Nonetheless, I wish that the default belief was that you *should* be able to leave your property unguarded, and that city life is, in this sense, a falling off from ideal standards. To institutionalise that idea that the default is that anything not locked or tied down is "fair game" is to bring in a grimmer society, in my opinion.
The function of bail is to make it not worth the accused's while skipping town. The problem for very poor people is that if they have nothing, there is no reason for them not to skip town. Therefore it is quite likely that even moderate bail will be set to more than their net worth. For a rich man, a level of bail which he can afford (possibly just by depositing securities) may still, in the judgement of the court, be such that the cost of skipping town exceeds the cost of facing the charges in court.
It is not so much an intentional "letting the rich get away with it". It is the problem that if you don't physically lock them up, you have no hold over the very poor, or over the fairly poor for serious offences. If you can find another way of ensuring that someone who has no assets will turn up in court, I am sure "the system" would love to know.
The problem with most "hardware" raid controllers is that they are actually software raid controllers: there is just a fairly conventional CPU on the Raid card doing exactly the same as the main CPU would do. Not many actually calculate the parity in dedicated hardware. The ones that do will give much better performance.
If you read further down, it says you can do a global reset, which loses the key and unlocks the disk as full of encrypted garbage, "with a few keystrokes".
That would require building a complete backup network of cell controllers, which cost the existing cell companies billions. iPhone certainly uses standard cell frequencies quick Google shows that Blackberry does in some places at least. Using new spectrum would also quite likely require designing different chips: the current cell frequencies are hard-engineered into the chips. In other words, anybody who wants real mobile access more or less has to use the existing infrastructure.
Or, of course, use much more bulky sat-phones, which I am sure that at least one or two of Obama's aides will be carrying - just in case. But those won't be Blackberries, they will be dumb phones.
The don't broadcast their GPS location, but they are in continuous communication with several cell controllers so they can hand over seamlessly as you move from cell to cell. And as they do so, they automatically regulate their transmit power level so that it only just reaches the controller, in order to minimise spill-over to other cells and hence minimise the number of bufer cells before you can re-use the same frequencies.
And, since power is inversely proportion to the square of distance, that actually means the cell controller knows how far away you are. Three such controllers, you can triangulate and find out where you are. And this happens all the time.
Used by the police in the UK when two girls were murdered. One of their mobile phones was last switched off outside the murderers house.
IBM is a pretty diverse company. Bits of it can be doing well while other bits are nose-diving. It could well be that the mainframe division (remember them?) is still churning our regular profits while the consultancy division has nose-dived. But you cannot recycle consultants into mainframe hardware engineers, so the consultants get the axe, even while the mainframe division is trying to re-invent the press gang. Or, of course, vice versa.
If you are going to make significant layoffs, it makes sense to do them at one time: get one headline, one dose of bad news, and get it out of the way. So I would guess that all the divisions will be instructed to get their worst case number, sum the lot, and announce that. No-one is going to complain if you come back and say "turns out it wasn't that bad - half the layoffs are cancelled", whereas if you have several rounds of layoffs moral will nosedive.
In normal times, announcing large layoffs may make the people you want to keep jittery and inclined to look elsewhere. But if there is enough generalised gloom, they probably won't.
So thinks may not be quite as black as the figures quoted at the head. Though, of course, those figures contribute to the general depression and make it worse. More positive feedback.
It may be to prevent wear, but it is more likely either SMART data gathering, as has been said, or a recalibrate to compensate for thermal changes. Seagate drives have a schedule for doing a recalibrate whether you like it or not, which plays hell with AV applications. But if you didn't do a recalibrate every now and again, differential thermal expansion means that the next write could go up to several tracks out from where you intended, losing the data written and destroying something else. If you got your update, storing data on that drive would become a lottery.
Or on the basis that it was, in the past, reasonable, and no-one is going to take the responsibility for rescinding an order that was sensible when it was given, and whose coses have already been accepted into the general cost of secure working.
Re:Google was just trying to save money
on
Google Router Rumors
·
· Score: 3, Informative
This sort of thing doesn't get offered, it is thrown in or dragged out as a sweetener for a humungeous order. And it is usually covered by a confidentiality clause because they don't want to be forced to offer it to the next, merely large, customer. But if you are placing an order which represents a serious fraction of quarter's output, you can get a lot thrown in - espexially if it doesn't actually cost anything to provide.
Though this would be a problem rather than a benefit for Google. They would have to put up fairly strong Chinese Walls inside their labs to ensure that the team developing their own router hadn't seen the competing device so couldn't be accused of ripping it off.
Subversion and git are not really comparable. Subversion is centralised, based on a single server for the repository, and therefore suitable for corporate use. git is decentralised and distributed and therefore more suitable for OSS with distributed developers and no or weak hierarchy, The competitors to git would be Bazaar and Mercurial.
Somebody in my company checked out all three and said that git was the most powerful, but also the most complex, hence a steep learning curve, and had poor Windows support. But presumably Perl developers are not frightened of complex software and unlikely to be using Windows as a primary platform, so git might well be appropriate for them. Bazaar was complemented for being Subversion like and Windows friendly (relevant to my company).
Anti-trust law would make it *very* difficult for the four parts of MS to co-operate. And a lot of code would need to be common (filesystems, the basic GUI, the kernel).
As to the second suggestion: what vehicle for nuisance litigation. You are doing a reasonable business as the only supplier of a relatively niche product. I come along and threaten to start up a competitor and force you to split - unless you buy me off. If you are making profits, very difficult for you to prove that the market would not be improved by my presence - so now the market must be split three ways. Who owns the patents? Who gets the guy who headed the design team and knows "everything"? Who gets the valuable brand name? Who gets to keep the factory in which the one machine that makes these machines lives?
One problem with this is to define a market. All vehicles, or just passenger cars, or just "traditional American gas guzzlers". All aircraft, or only 400+ seat aircraft (i.e. Boeing's thirty year monopoly with the 747)? If you introduce a brand new product (the microprocessor, USB sticks, commercial orbital launchers, a new drug) then by definition you own 100% of the market the day you sell your first product. Windows has 90% of the market for desktop OSes. Would it do any good to split MS up into four different suppliers of the identical code? Or to have four diverging implementations of Windows because the four companies would be prevented from co-operating by anti-trust laws. Splitting off, say, Office wouldn't reduce the OS competition.
The problem comes with the concept of "too big to fail" due to concentration. In the general case, you are quite right, If you have a hundred bakers, and there is not enough demand for bread to keep them all in business, the least efficient ones go bankrupt and the more efficient ones thrive, thus increasing the general efficiency of the baking business.
The problem is when you have monolithic industries, or quasi monolithic industries, where the whole industry for one region sinks together. If you have one mega-bakery and that is badly run, you cannot let it go bust because there will be no bread and people will starve.
The DRAM industry has become increasingly capital intensive: a new Fab costs billions, but produces vast numbers of chips. This means that there are very few fabs in one country, probably only one. Healthy shrinkage, by a few percent, is not possible: you lose 50% of your industry or none. And, politically, that is unacceptable for any single country. Worldwide, it would be correct to close down one or two DRAM fabs at the moment. A World Government, with perhaps 12 fabs, could look at the big picture. But Taiwan, Germany, and South Korea, with two or three each, cannot accept losing such a large slice of their industry.
The same was true in the finance industry: because of their size, AIG, Fanny Mae and Freddy Mac were "too big to fail". Lehmann was adjudge not to be too big - but the repercussions of its failure are turning out much larger than expected.
The same is true in cars. "Detroit", the Big Three American car manufacturers, is collectively "too big to fail". And they are so interlocked in the public mind that they would appear to sink or swim together. Mind you, their problems are basically the result of baling out Chrysler twice instead of letting it fail at a time when it could have failed on its own and brought the appropriate slimming down to Detroit.
Slashdot Mirror
So if we extrapolate this trend, nuclear fusion power is only 30 years away!
Always has been, always will be.
Cash is not necessarily cheaper for the merchant. The cost of secure transport of cash to the bank can be the same as merchant fees. On reason some supermarkets offer cashback - it reduces their cash transport costs.
No, that is what Hawking, and a considerable number of other scientists believe. Essentially, nature is allowed to "borrow" energy from nowhere provided the product of the energy and time the energy exists does not exceed Planks constant. When it does so, a particle and its matching antiparticle (to keep all the charges, baryon numbers etc. matched) spring into existence for a very short time, then cancel out again, "repaying" the borrowed energy.
Except that if this happens really close to the event horizon of a black hole, one of the two particles can fall into the hole and the other doesn't, resulting in the net creation of a particle outside the event horizon. The energy needed to "balance the books" and create the particle comes from the black hole. This means that black holes are continuously emitting particles, which are called Hawking Radiation, and losing energy. However, to maximise the chance of one particle falling in and the other escaping, the gravitational field has to be very non-linear, which means that the hole has to be small. The smaller the hole, the faster it evaporates, so the faster it shrinks which eventually leads to a runaway; tiny black holes explode. However, stellar mass black holes evaporate so slowly that it takes a bucket load of exponents to measure the time until they explode.
I have used it for successfully to get out at the right stop when I cannot read the station names. You get out when the train is scheduled to arrive at your destination. My experience is that (a) the train is actually stopped within the scheduled minute, and (b) it is at the right destination. Very comforting when the script it complete gibberish to you.
This is probably a problem in jobs where people want interchangeable droids. But the sort of jobs wanted by the sort of people who read /. want people with originality and flexibility. And if you are original and flexible, you are probably going to have an original and flexible private life. So if an employer is going to get upset about the somewhat odd (but legal) things you may get up to at home, they are probably so strait-laced that you don't want to work for them, and they are probably not going to succeed in the long term. (If you do illegal things AND post them to the Net, you are too stupid to deserve the job.)
Certainly a net search on my name reveals a lot about me - very boring, mostly, but accurate. But if any employer objects to any of the trail I have left, I probably don't want to work for them anyway. Of course, when the economy crumbles to the point I am reduced to burger flipping, I may change my mind.
Also, that is machine code, not SQL. SQL has an built-in date type, whose internal representation is opaque, with which it is capable of doing date arithmetic e.g. finding the number of days between two dates expressed, in human terms, as year/month/day. One of the points of using a proper database it to avoid having to re-invent all the permutations of date arithmetic.
You could perfectly well allocate magic dates for special meanings; this would be regarded by most database professionals as the dirtiest of kludges.
I said not that it means anything in SQL. I agree that it means "absence of data". But a database designer must specify what "absence of data" means in the context of this table, just as a C programmer must know what "not pointing to anything" (aka a C NULL) means in his context. Does "absence of data" in the Date of Death column mean the person is alive, or that the database does not know his death date? A matter of design, not programming.
How much memory will you pay to avoid one bug? Null terminated strings were invented when 64k was a large memory for a single computer. We have moved rather a long way from that. I would spend a *lot* of ram to save one bug in delivered code.
Which is solved if lists and trees are implemented within the system in the same way as integers and floats. Lists are implemented in Python. The article was referring to language design: I think most language designers today would ensure that lists, both fixed and variable, were either in the language or easily implemented in efficient library functions. Trees don't seem to have the same near-universal acceptance.
What NULL "means" is undefined. There are at least two possible meanings, and it is up to the database designer to define, for every column in which null is allowed. The two meanings obvious meanings are either "unknown" or "has not got". For a Date Of Birth column, it is obviously "unknown", because everybody has a DOB. But what does a null Date Of Death column mean? In a commercial context, it probably means that, so far as the database is concerned, the person is still alive. But in a historical database, it could mean that the DOD is really unknown, or that the person is alive. Likewise an Address3 line might interpret NULL as Address3 is "unknown" if Address1 is null and "has not got" if Address1 is non-null. Good database design will always specify this.
There are some places - the Scottish Islands are one - where the crime rate is low enough that people routinely leave their houses unlocked. Neighbours can enter the house, e.g. to borrow and return things, at will. Likewise car keys are left in the ignition so that if the car is in the way anybody can move it.
I realise that it is impossibly idealistic to expect this to work in cities. Nonetheless, I wish that the default belief was that you *should* be able to leave your property unguarded, and that city life is, in this sense, a falling off from ideal standards. To institutionalise that idea that the default is that anything not locked or tied down is "fair game" is to bring in a grimmer society, in my opinion.
The function of bail is to make it not worth the accused's while skipping town. The problem for very poor people is that if they have nothing, there is no reason for them not to skip town. Therefore it is quite likely that even moderate bail will be set to more than their net worth. For a rich man, a level of bail which he can afford (possibly just by depositing securities) may still, in the judgement of the court, be such that the cost of skipping town exceeds the cost of facing the charges in court.
It is not so much an intentional "letting the rich get away with it". It is the problem that if you don't physically lock them up, you have no hold over the very poor, or over the fairly poor for serious offences. If you can find another way of ensuring that someone who has no assets will turn up in court, I am sure "the system" would love to know.
The problem with most "hardware" raid controllers is that they are actually software raid controllers: there is just a fairly conventional CPU on the Raid card doing exactly the same as the main CPU would do. Not many actually calculate the parity in dedicated hardware. The ones that do will give much better performance.
If you read further down, it says you can do a global reset, which loses the key and unlocks the disk as full of encrypted garbage, "with a few keystrokes".
That would require building a complete backup network of cell controllers, which cost the existing cell companies billions. iPhone certainly uses standard cell frequencies quick Google shows that Blackberry does in some places at least. Using new spectrum would also quite likely require designing different chips: the current cell frequencies are hard-engineered into the chips. In other words, anybody who wants real mobile access more or less has to use the existing infrastructure.
Or, of course, use much more bulky sat-phones, which I am sure that at least one or two of Obama's aides will be carrying - just in case. But those won't be Blackberries, they will be dumb phones.
The don't broadcast their GPS location, but they are in continuous communication with several cell controllers so they can hand over seamlessly as you move from cell to cell. And as they do so, they automatically regulate their transmit power level so that it only just reaches the controller, in order to minimise spill-over to other cells and hence minimise the number of bufer cells before you can re-use the same frequencies.
And, since power is inversely proportion to the square of distance, that actually means the cell controller knows how far away you are. Three such controllers, you can triangulate and find out where you are. And this happens all the time.
Used by the police in the UK when two girls were murdered. One of their mobile phones was last switched off outside the murderers house.
IBM is a pretty diverse company. Bits of it can be doing well while other bits are nose-diving. It could well be that the mainframe division (remember them?) is still churning our regular profits while the consultancy division has nose-dived. But you cannot recycle consultants into mainframe hardware engineers, so the consultants get the axe, even while the mainframe division is trying to re-invent the press gang. Or, of course, vice versa.
If you are going to make significant layoffs, it makes sense to do them at one time: get one headline, one dose of bad news, and get it out of the way. So I would guess that all the divisions will be instructed to get their worst case number, sum the lot, and announce that. No-one is going to complain if you come back and say "turns out it wasn't that bad - half the layoffs are cancelled", whereas if you have several rounds of layoffs moral will nosedive.
In normal times, announcing large layoffs may make the people you want to keep jittery and inclined to look elsewhere. But if there is enough generalised gloom, they probably won't.
So thinks may not be quite as black as the figures quoted at the head. Though, of course, those figures contribute to the general depression and make it worse. More positive feedback.
It may be to prevent wear, but it is more likely either SMART data gathering, as has been said, or a recalibrate to compensate for thermal changes. Seagate drives have a schedule for doing a recalibrate whether you like it or not, which plays hell with AV applications. But if you didn't do a recalibrate every now and again, differential thermal expansion means that the next write could go up to several tracks out from where you intended, losing the data written and destroying something else. If you got your update, storing data on that drive would become a lottery.
The magnetic field recorded in lava flows has lasted hundreds of millions of years.
That's it! Put a coil round an active volcano and write your long term data into the lava. No more lost backups now!
Or on the basis that it was, in the past, reasonable, and no-one is going to take the responsibility for rescinding an order that was sensible when it was given, and whose coses have already been accepted into the general cost of secure working.
This sort of thing doesn't get offered, it is thrown in or dragged out as a sweetener for a humungeous order. And it is usually covered by a confidentiality clause because they don't want to be forced to offer it to the next, merely large, customer. But if you are placing an order which represents a serious fraction of quarter's output, you can get a lot thrown in - espexially if it doesn't actually cost anything to provide.
Though this would be a problem rather than a benefit for Google. They would have to put up fairly strong Chinese Walls inside their labs to ensure that the team developing their own router hadn't seen the competing device so couldn't be accused of ripping it off.
Subversion and git are not really comparable. Subversion is centralised, based on a single server for the repository, and therefore suitable for corporate use. git is decentralised and distributed and therefore more suitable for OSS with distributed developers and no or weak hierarchy, The competitors to git would be Bazaar and Mercurial.
Somebody in my company checked out all three and said that git was the most powerful, but also the most complex, hence a steep learning curve, and had poor Windows support. But presumably Perl developers are not frightened of complex software and unlikely to be using Windows as a primary platform, so git might well be appropriate for them. Bazaar was complemented for being Subversion like and Windows friendly (relevant to my company).
Anti-trust law would make it *very* difficult for the four parts of MS to co-operate. And a lot of code would need to be common (filesystems, the basic GUI, the kernel).
As to the second suggestion: what vehicle for nuisance litigation. You are doing a reasonable business as the only supplier of a relatively niche product. I come along and threaten to start up a competitor and force you to split - unless you buy me off. If you are making profits, very difficult for you to prove that the market would not be improved by my presence - so now the market must be split three ways. Who owns the patents? Who gets the guy who headed the design team and knows "everything"? Who gets the valuable brand name? Who gets to keep the factory in which the one machine that makes these machines lives?
One problem with this is to define a market. All vehicles, or just passenger cars, or just "traditional American gas guzzlers". All aircraft, or only 400+ seat aircraft (i.e. Boeing's thirty year monopoly with the 747)? If you introduce a brand new product (the microprocessor, USB sticks, commercial orbital launchers, a new drug) then by definition you own 100% of the market the day you sell your first product. Windows has 90% of the market for desktop OSes. Would it do any good to split MS up into four different suppliers of the identical code? Or to have four diverging implementations of Windows because the four companies would be prevented from co-operating by anti-trust laws. Splitting off, say, Office wouldn't reduce the OS competition.
The problem comes with the concept of "too big to fail" due to concentration. In the general case, you are quite right, If you have a hundred bakers, and there is not enough demand for bread to keep them all in business, the least efficient ones go bankrupt and the more efficient ones thrive, thus increasing the general efficiency of the baking business.
The problem is when you have monolithic industries, or quasi monolithic industries, where the whole industry for one region sinks together. If you have one mega-bakery and that is badly run, you cannot let it go bust because there will be no bread and people will starve.
The DRAM industry has become increasingly capital intensive: a new Fab costs billions, but produces vast numbers of chips. This means that there are very few fabs in one country, probably only one. Healthy shrinkage, by a few percent, is not possible: you lose 50% of your industry or none. And, politically, that is unacceptable for any single country. Worldwide, it would be correct to close down one or two DRAM fabs at the moment. A World Government, with perhaps 12 fabs, could look at the big picture. But Taiwan, Germany, and South Korea, with two or three each, cannot accept losing such a large slice of their industry.
The same was true in the finance industry: because of their size, AIG, Fanny Mae and Freddy Mac were "too big to fail". Lehmann was adjudge not to be too big - but the repercussions of its failure are turning out much larger than expected.
The same is true in cars. "Detroit", the Big Three American car manufacturers, is collectively "too big to fail". And they are so interlocked in the public mind that they would appear to sink or swim together. Mind you, their problems are basically the result of baling out Chrysler twice instead of letting it fail at a time when it could have failed on its own and brought the appropriate slimming down to Detroit.