No. Delimiting "who set up... was trusted" with commas marks it as a subsidiary clause. It isn't a subsidiary clause -- the meaning of the sentence would be completely changed without it.
Commas should be used, like this, to delimit a clause only when that clause is less important than the rest of the sentence.
What is that sentence supposed to mean? Use a freaking comma!
Where would you put a comma in that sentence? Commas do not exist simply for the purpose of being scattered randomly.
The only correction necessary would be to remove the extraneous "to": It discusses a case in which an administrator who set up back doors in the system with which he was trusted deleted files to which he could access after he was fired.
What if you are on unemployment, and you get only one job offer, and that offer includes a clause that requires you to agree not to read Slashdot. And because it is an offer, if you turn it down, you could lose your unemployment benefits.
If you're really worried about that possibility, sign a contract with someone which requires you to read slashdot.
That way, when you get that job offer, you don't have to report it because you weren't legally able to accept it.
Personally, I think the employers with whom you have a non-compete agreement should be the ones paying you unemployment benefits
They may not be paying unemployment benefits, but they *do* provide compensation for the non-compete agreement. In the case of slave traders like these, the compensation is in the form of getting a job in the first place; in the case of other companies, people signing non-compete agreements are generally paid more than they would receive at a job which did not require such an agreement.
If you don't like the terms of employment offered, *don't accept them*.
GPUs are effectively stream processors, a class of devices whose architecture and programming model make then particularly efficient for scientific calculation.
Unfortunately, the "128 bit colour" is just four 32-bit elements (RGB+alpha). If were actually using 128 bit floating point arithmetic, they'd call it "512 bit colour".;)
GPUs work with limited precision -- IEEE single precision is typical. This is good enough for 3D graphics -- after all, in the end you'll be limited by the 10-11 bit spatial resolution and 8 bit color resolution -- but not good enough for most scientific problems, which typically require a minimum of double precision.
Simulating higher precision with single precision arithmetic is possible, but the performance penalty is too severe for it to be useful.
A license is a contract. You can't enter into a contract by mistake -- or rather, if you do, the courts can overrule it.
Providing that SCO can show evidence that they made a "good faith" attempt to stop distributing linux-plus-their-proprietary-code as soon as possible, the inadvertant GPLing of their code would be overturned by the courts.
Given that it would be illegal to email such an image, this is something you'd only do if the email were truly anonymous.
Distributing child pornography for the purpose of reporting it to law enforcement is not illegal. (Even if it violates a literal reading of the law, no jury would ever convict.)
If you think that it is no longer possible for young people to make significant contributions to mathematics, just look at the various fields medal winners of the past decade.
I was looking forward to a hypothetical future where working out the structure of a folded protein is easy, given the nucleotide sequence, but constructing a sequence which will result in a given structure is harder. I can imagine that the "programmers" who would construct such sequences would be very much like early assembly language programmers.
A century ago, mathematics was primarily a new field. New fields are characterized by inventiveness and a lack of prerequisite knowledge -- there isn't a lot of background to learn, and if you look at problems "the right way" you can get results very quickly. Most of mathematics is no longer a new field; in most areas, one must spend years studying before one can do anything new, and even then it's likely to be the result of long hard work rather than a quick new insight.
Computer science is moving in the same direction, but is many years behind. Thirty years ago, computer science was a new field; there were few if any courses teaching necessary background material; and someone with the right insight could find very important work very easily. Now, we're starting to see movement away from that -- there is a body of important work to build upon, and anyone who hasn't studied that work will have "new insights" which simply reinvent already existing work.
Mathematics is no longer a young man's game, and this is probably the last generation when computer science has been a young man's game. Next generation, the young will find a new field to excel in -- perhaps genomics?
Slashdot Mirror
I'd say that the BBC has more in common with the Normans, actually.
No. Delimiting "who set up ... was trusted" with commas marks it as a subsidiary clause. It isn't a subsidiary clause -- the meaning of the sentence would be completely changed without it.
Commas should be used, like this, to delimit a clause only when that clause is less important than the rest of the sentence.
What is that sentence supposed to mean? Use a freaking comma!
Where would you put a comma in that sentence? Commas do not exist simply for the purpose of being scattered randomly.
The only correction necessary would be to remove the extraneous "to":
It discusses a case in which an administrator who set up back doors in the system with which he was trusted deleted files to which he could access after he was fired.
What if you are on unemployment, and you get only one job offer, and that offer includes a clause that requires you to agree not to read Slashdot. And because it is an offer, if you turn it down, you could lose your unemployment benefits.
If you're really worried about that possibility, sign a contract with someone which requires you to read slashdot.
That way, when you get that job offer, you don't have to report it because you weren't legally able to accept it.
Personally, I think the employers with whom you have a non-compete agreement should be the ones paying you unemployment benefits
They may not be paying unemployment benefits, but they *do* provide compensation for the non-compete agreement. In the case of slave traders like these, the compensation is in the form of getting a job in the first place; in the case of other companies, people signing non-compete agreements are generally paid more than they would receive at a job which did not require such an agreement.
If you don't like the terms of employment offered, *don't accept them*.
We're talking about C++ here, not perl.
Compiled C++ code can't be decompiled into anything approximating the readability of the original; compiled perl code can.
It's called "evaporative cooling".
The article refers to this as a red dwarf, but also describes it as being 0.07 solar masses. Is it possible for a red dwarf to be that lightweight?
I think that's the first time I've ever seen Microsoft compared to Communism.
Usually the comparison is the other way around.
Quite a few bacteria have travelled to the moon.
Out of curiosity, what is a "health salt"?
You might have heard of it under a different name: Snake oil.
GPUs are effectively stream processors, a class of devices whose architecture and programming model make then particularly efficient for scientific calculation.
Also known as vector processors.
The more things change...
Unfortunately, the "128 bit colour" is just four 32-bit elements (RGB+alpha). If were actually using 128 bit floating point arithmetic, they'd call it "512 bit colour". ;)
"128-bit" colour precision is just a four dimensional vector of 32-bit elements.
GPUs work with limited precision -- IEEE single precision is typical. This is good enough for 3D graphics -- after all, in the end you'll be limited by the 10-11 bit spatial resolution and 8 bit color resolution -- but not good enough for most scientific problems, which typically require a minimum of double precision.
Simulating higher precision with single precision arithmetic is possible, but the performance penalty is too severe for it to be useful.
A license is a contract. You can't enter into a contract by mistake -- or rather, if you do, the courts can overrule it.
Providing that SCO can show evidence that they made a "good faith" attempt to stop distributing linux-plus-their-proprietary-code as soon as possible, the inadvertant GPLing of their code would be overturned by the courts.
"Java Performance Urban Legends" should read "Java Performance is an Urban Legend".
Looking for child pornography for "research purposes" is quite a different matter from stumbling across it by accident and storing the evidence.
Given that it would be illegal to email such an image, this is something you'd only do if the email were truly anonymous.
Distributing child pornography for the purpose of reporting it to law enforcement is not illegal. (Even if it violates a literal reading of the law, no jury would ever convict.)
Well, the catch is that when the police don't find anything, they'd investigate where the "anonymous" message came from.
If you think that it is no longer possible for young people to make significant contributions to mathematics, just look at the various fields medal winners of the past decade.
Yeah, it's really amazing... all the people who win a prize which is only awarded to people at most 40 years old is never awarded to anyone over 40 years old.
Obviously, one would include some evidence. Send an email which reads "I found this on Foo Bar's computer" and attach one of the images.
Sure, someone could be falsely accused in this manner, but the accuser would need to have gained access to the picture from somewhere.
I was looking forward to a hypothetical future where working out the structure of a folded protein is easy, given the nucleotide sequence, but constructing a sequence which will result in a given structure is harder. I can imagine that the "programmers" who would construct such sequences would be very much like early assembly language programmers.
A century ago, mathematics was primarily a new field. New fields are characterized by inventiveness and a lack of prerequisite knowledge -- there isn't a lot of background to learn, and if you look at problems "the right way" you can get results very quickly. Most of mathematics is no longer a new field; in most areas, one must spend years studying before one can do anything new, and even then it's likely to be the result of long hard work rather than a quick new insight.
Computer science is moving in the same direction, but is many years behind. Thirty years ago, computer science was a new field; there were few if any courses teaching necessary background material; and someone with the right insight could find very important work very easily. Now, we're starting to see movement away from that -- there is a body of important work to build upon, and anyone who hasn't studied that work will have "new insights" which simply reinvent already existing work.
Mathematics is no longer a young man's game, and this is probably the last generation when computer science has been a young man's game. Next generation, the young will find a new field to excel in -- perhaps genomics?
When will people learn that if you're going to download program updates, you should use public-key cryptography to sign the updates?
If you're going to write a worm, do it right.