Some of the cited advantages of not using garbage collection are red herrings. For instance the "controlling access to files and similar resources" by RAII works fine with garbage collection. In most cases, the compiler can determine by static analysis that a particular object is allocated within a scope and no referenes are propogated upward out of scope, and can remove the reference so the garbage collector will deallocate it (possibly calling a destructor). Depending on the type of GC and its implementation, the compiler may generate code that forces the object to be deallocated immediately.
For cases where static analysis can't do this automatically, it isn't that hard to use a design methodology that achieves the same result; it's certainly still much easier than doing manual allocation and deallocation and ensuring that the deallocation is done (or not done) correctly in all cases.
And if you are using a reference-counting GC, or a hybrid GC that includes reference-counting, you don't have to do anything special at all.
The same applies to the claimed mutex and error message disadvantages, since those are just specific uses of RAII.
Actually, every time I've voted with a paper ballot (not punched card) in California, I have in fact gotten a receipt. The receipt doesn't show my votes, though; it just has the serial number of the ballot.
Thanks for the info. It looks like they only offer Masters degrees in CS and Software
Engineering by distance learning, not Bachelor's degrees. But I'll research it further.
The other two I'm looking at based on Slashdot comments are UIS (University of Illinois at Springfield) and University of Hawaii.
I've been researching getting a BSc in Computer Science online for a few weeks. I've been employed as a programmer for embedded systems (mostly C and assembly) for 19 years, make a good salary, and having a BSc isn't likely to get me a better salary, promotion, etc. My interest in getting a BSc is mainly so that I can get into a Master's program (not necessarily in CS), which I want to do primarily for my own personal satisfaction.
I did a Google search and started looking into various programs. I was able to immediately reject anything in "Information Technology". But it appears that a lot of the online degrees called CS aren't really what I'd consider CS. Many of them seem to be misnamed degrees in programming. You take a few semesters of C, a few of C++, and a few of Java, and they give you a CS degree. As far as I'm concerned, that's complete bullshit. Sure, CS requires knowledge of programming, but it goes far beyond that. IMNSHO, a CS degree should also require course work in discrete structures, compiler construction, complexity theory, computer organization, and perhaps some specialties like graphics, robotics, error control coding, etc.
If I were hiring a recent CS grad (whether online or bricks and mortar), and saw that their coursework was mostly just classes in programming in various languages, I wouldn't necessarily reject the candidate, but I'd certainly be very wary.
This leads to the big question: What accredited universities offer a "Real" BSc in CS online?
My own sugestion is to drill a hole into the ground as far as is possible i.e. several kilometres, let off an appropriate nuke to create an underground chamber.
I'm sure the rest of the world would enjoy hearing about such a violation of the Test Ban Treaty.
It doesn't violate the test ban treaty because it isn't a test. It's production use for excavation.
Sure, and last year all of the DARPA Grand Challenge entries made it "not quite" to the finish line. But just as there were more successful entries this year, there will be more successful space elevator technology demonstrations in the future.
The article seems to try to imply that there's something wrong with having to apply "historical legal theories" to "Cyberspace", as if they're so antiquated as to be mere curiosities or something. But this looks to me like a cut-and-dried case where applying the standard legal doctrine achieves an entirely reasonable outcome, and shows that at least some of the "historical legal theories" are in fact protecting people in exactly the same way today as they did in the past, despite the advance of technology.
Plus, there is some strategy here too. Sometimes in chess, you'll let a pawn be sacrificed to set up a checkmate.
Yes, you sacrifice a pawn when it improves your position in other areas. Dropping the patent counterclaims doesn't improve IBM's position, so this analogy is not applicable.
Your writer friend is getting paid for the right to publish her work.
Most people writing GPL'd software don't get paid anything for it. I've written various fairly obscure GPL'd software, and aside from some donations from users that are happy with it, I don't get paid for it at all. I certainly am not going to indemnify anyone against damages.
Ask your writer friend if she would provide indemnification if she did not get paid for the publishing rights, and only got a few donations from readers now and then.
If someone wants indemnification when using GPL'd software, they can negotiate a contract with the author. For example, I think I could probably provide $100,000 of indemnification to a customer that is willing to pay me $2000, though I'd obviously have to check with my insurance carrier for their exact details and rates before I'd sign such a contract.
A green little chemist,
On a green little day,
Mixed some green little chemicals
In a green little way.
The green little grasses
Now tenderly wave
Over the green little chemist's
Green little grave.
The secret to enjoying your job...
on
Pay vs. Happiness
·
· Score: 2, Insightful
...is to work in a field that you're personally interested in, for a company that is small enough that they do care about the employees. You'll rarely, if ever, find that in a large company; HP was once that way but Carly "fixed" that.
Most large companies consider employees to be completely interchangeable and replaceable like light bulbs.
That's not to say that all small companies are good, though. Many tech startups have a business plan that requires making their employees work long hours and weekends until they burn out. Avoid those like the plauge. They always tell prospective employees that they will reap big rewards on stock options, and in fact often insist that the employee should accept lower salary and worse benefits in exchange for the options. Don't buy it. Options *might* pay off, but it's a long shot. If they try to sucker you into such a plan, ask them to give you the salary and benefits you want and forget the options. They'll almost never do that, which tells you that their real opinion on the value of their own options is that they are worthless; obviously you shouldn't value them any more highly.
I've had the good fortune to have several enjoyable jobs at small companies, including my current job. At a few of them I did eventually make modest gains on stock options, but not enough for a down payment on a house. Well, maybe a down payment on a house somewhere other than in Silicon Valley.
And what I'm saying is that the president does NOT have the power to wage a war without a declaration from Congress, because doing so effectively constitutes a declaration of war.
The fact that many presidents have done it does not prove that it is legitimate. The government does lots of things that they aren't constituionally authorized to do.
Over a couple centuries there has been significant legislation to limit the president's powers.
That's irrelevant. It is meaningless for Congress to pass an act telling the president that he's not allowed to do something that the constitution doesn't grant him power to do in the first place.
The constitution grants Congress the power to declare war. NOT the president. And Congress does not have the power to delegate that authority to declare war to the president. This has nothing to do with the War Powers Act.
So unless you are going to argue that dropping a nuke on another country as a first strike is not a declaration of war, it is clear that the president does not have the authority to do so.
As I pointed out originally, the fact that Congress or the president don't legally have the power to do something rarely if ever actually prevents them from doing it.
Linux is not UNIX, and I believe all UNIX code was removed from BSD as well, so those are not UNIX schedulers.
They aren't AT&T code, but they are UNIX. UNIX is a trademark of The Open Group, and does not refer to any specific source code. Neither Linux nor FreeBSD are certified to comply with the Single Unix Standard, but both are at least mostly compliant. I suppose if you want to be picky, only AIX and Solaris are certified as "Unix 03", but I doubt that anyone reasonable person reading my posting would have assumed that I meant my usage of "Unix" to be that restrictive.
First, torrents are neither CPU nor disk bound.
Even if they are network-bound, they still shouldn't interfere with scheduling of non-network processes. But in my experience, they do.
Second, I don't observe this behavior on my machines, and my Internet connection is very fast.
I don't know what "very fast" is, but my internet connection is 100Mbps (peak).
I have yet to see a better scheduler than the UNIX schedulers;
The TOPS-10 scheduler on the DECsystem-10 was far better than any Unix scheduler I've ever seen. Modern Unix schedulers are written by people that never had to deal with a real multiuser timesharing system supporting varied workloads while still remaining responsive for interactive use. Nowdays it's assumed that the system is effectively supporting a single user, without considering that the single user would still want good interactive response even while he or she has some CPU-intensive and/or disk-intensive background processes running.
As an example, on an Athlon XP 2500 system with a 3ware RAID controller, running BitTorrent with only a few torrents will completely hose interactive response with Linux 2.6 and FreeBSD 4 kernels. (I haven't yet tried FreeBSD 5.)
Perhaps I wasn't clear enough. My point was that the president doesn't have authority to grant himself (or the Pentagon) the power to launch a preemptive nuclear strike.
I don't see how they think they have the authority to let the president
authorize a first strike. The power to declare
war belongs to the Congress, not the president, and the War Powers
Resolution of 1973 limits the power of the President of the United States
to wage war without the approval of the Congress.
Of course, since W's administration doesn't seem to think the
Constitution is worth the paper it's printed on, this won't stop them.
And Congress doesn't seem to hold it in any higher regard these days.
The Constitution says that Congress has the power to declare war, not
the power to issue an "authorization of force".
Compare a Palm to a Sharp Zaurus, and you will be disgusted to even touch the Palm again.
I have a Zaurus and several Palms. As a Linux user/programmer, I do have a lot of fun with the Zaurus. But the usability of the Zaurus as a PDA is terrible, mostly because Qtopia is incredibly buggy and the UI is poorly designed.
PalmOS is not as powerful, but it is much more robust and has a much better UI.
Regarding your car argument, yes, whatever contract I sign regarding how I use the car is fully enforceable, however such contracts don't exist with car purchases.
That's exactly my point. When I purchased a new car a year ago, there was nothing in the fine print that said that I was licensing GM's patents. And yet GM still cannot (successfully) sue me for patent infringement for owning or using the car, because the patent rights were exhausted for that new car when I purchased it (first sale).
In a similar case, Jazz Photo Corp v International Trade Commission, it was determined that the item was not licensed because the purchaser had insufficient foreknowledge that such a license was an issue.
And in the Lexmark case, the purchaser similarly had no foreknowledge that he or she was (as Lexmark claims) not permitted to use Lexmarks patents with regard to that cartridge.
In the Lexmark case, it's clear that this is licensed - the purchaser fills out a form and sends it to Lexmark agreeing that they will not refill it
It's a contract of adhesion, not a license. The statement printed on the box that the purchaser agrees to send the cartridge to Lexmark when he or she has finished using it does not indicate that the cartridge was not sold, or that it does not belong to the purchaser. In fact, it didn't even say that the user could not refill it; the Ninth Cirucit (improperly, IMNSHO) decided that the words "single use" found elsewhere on the packaging were part of the contract.
If Lexmark's lawyers had been on the ball, they would have specifically inserted wording that the "purchaser" was only purchasing a right to use the cartrige, and not ownership, though it is not clear that even such language in a contract of adhesion would normally be considered enforceable. For example, see Softman v. Adobe (2001), in which it was ruled that a transaction conducted in the manner of a sale was legally a sale, despite the seller's efforts to construe it as a license.
And in any case, the agreement (whether a contract or a license) does not say anything about patents - specifically it does not say that the item is not licensed under Lexmark's patents. If the purchaser fails to abide by the contract terms, Lexmark's recourse is to sue the purchaser for breach of contract, not for patent infringement. At least, that was the legal situation before this Ninth Circuit ruling.
Slashdot Mirror
A live person is missing a face. A dead person doesn't need theirs any more. Where's the problem?
And how could the "psychological impact" be worse than not havin a face? The patient is going to "look different" no matter what is done.
For cases where static analysis can't do this automatically, it isn't that hard to use a design methodology that achieves the same result; it's certainly still much easier than doing manual allocation and deallocation and ensuring that the deallocation is done (or not done) correctly in all cases.
And if you are using a reference-counting GC, or a hybrid GC that includes reference-counting, you don't have to do anything special at all.
The same applies to the claimed mutex and error message disadvantages, since those are just specific uses of RAII.
Actually, every time I've voted with a paper ballot (not punched card) in California, I have in fact gotten a receipt. The receipt doesn't show my votes, though; it just has the serial number of the ballot.
The other two I'm looking at based on Slashdot comments are UIS (University of Illinois at Springfield) and University of Hawaii.
I did a Google search and started looking into various programs. I was able to immediately reject anything in "Information Technology". But it appears that a lot of the online degrees called CS aren't really what I'd consider CS. Many of them seem to be misnamed degrees in programming. You take a few semesters of C, a few of C++, and a few of Java, and they give you a CS degree. As far as I'm concerned, that's complete bullshit. Sure, CS requires knowledge of programming, but it goes far beyond that. IMNSHO, a CS degree should also require course work in discrete structures, compiler construction, complexity theory, computer organization, and perhaps some specialties like graphics, robotics, error control coding, etc.
If I were hiring a recent CS grad (whether online or bricks and mortar), and saw that their coursework was mostly just classes in programming in various languages, I wouldn't necessarily reject the candidate, but I'd certainly be very wary.
This leads to the big question: What accredited universities offer a "Real" BSc in CS online?
Sure, and last year all of the DARPA Grand Challenge entries made it "not quite" to the finish line. But just as there were more successful entries this year, there will be more successful space elevator technology demonstrations in the future.
I want malk (Now with vitamin R!),
...have a GOOD grammar checker?
The article seems to try to imply that there's something wrong with having to apply "historical legal theories" to "Cyberspace", as if they're so antiquated as to be mere curiosities or something. But this looks to me like a cut-and-dried case where applying the standard legal doctrine achieves an entirely reasonable outcome, and shows that at least some of the "historical legal theories" are in fact protecting people in exactly the same way today as they did in the past, despite the advance of technology.
Most people writing GPL'd software don't get paid anything for it. I've written various fairly obscure GPL'd software, and aside from some donations from users that are happy with it, I don't get paid for it at all. I certainly am not going to indemnify anyone against damages.
Ask your writer friend if she would provide indemnification if she did not get paid for the publishing rights, and only got a few donations from readers now and then.
If someone wants indemnification when using GPL'd software, they can negotiate a contract with the author. For example, I think I could probably provide $100,000 of indemnification to a customer that is willing to pay me $2000, though I'd obviously have to check with my insurance carrier for their exact details and rates before I'd sign such a contract.
Most large companies consider employees to be completely interchangeable and replaceable like light bulbs.
That's not to say that all small companies are good, though. Many tech startups have a business plan that requires making their employees work long hours and weekends until they burn out. Avoid those like the plauge. They always tell prospective employees that they will reap big rewards on stock options, and in fact often insist that the employee should accept lower salary and worse benefits in exchange for the options. Don't buy it. Options *might* pay off, but it's a long shot. If they try to sucker you into such a plan, ask them to give you the salary and benefits you want and forget the options. They'll almost never do that, which tells you that their real opinion on the value of their own options is that they are worthless; obviously you shouldn't value them any more highly.
I've had the good fortune to have several enjoyable jobs at small companies, including my current job. At a few of them I did eventually make modest gains on stock options, but not enough for a down payment on a house. Well, maybe a down payment on a house somewhere other than in Silicon Valley.
(There are monkeyboys in the facility.)
The fact that many presidents have done it does not prove that it is legitimate. The government does lots of things that they aren't constituionally authorized to do.
The constitution grants Congress the power to declare war. NOT the president. And Congress does not have the power to delegate that authority to declare war to the president. This has nothing to do with the War Powers Act.
So unless you are going to argue that dropping a nuke on another country as a first strike is not a declaration of war, it is clear that the president does not have the authority to do so.
As I pointed out originally, the fact that Congress or the president don't legally have the power to do something rarely if ever actually prevents them from doing it.
As an example, on an Athlon XP 2500 system with a 3ware RAID controller, running BitTorrent with only a few torrents will completely hose interactive response with Linux 2.6 and FreeBSD 4 kernels. (I haven't yet tried FreeBSD 5.)
Perhaps I wasn't clear enough. My point was that the president doesn't have authority to grant himself (or the Pentagon) the power to launch a preemptive nuclear strike.
And you think that a preemptive nuclear striked doesn't constitute a declaration of war?
Of course, since W's administration doesn't seem to think the Constitution is worth the paper it's printed on, this won't stop them.
And Congress doesn't seem to hold it in any higher regard these days. The Constitution says that Congress has the power to declare war, not the power to issue an "authorization of force".
PalmOS is not as powerful, but it is much more robust and has a much better UI.
If Lexmark's lawyers had been on the ball, they would have specifically inserted wording that the "purchaser" was only purchasing a right to use the cartrige, and not ownership, though it is not clear that even such language in a contract of adhesion would normally be considered enforceable. For example, see Softman v. Adobe (2001), in which it was ruled that a transaction conducted in the manner of a sale was legally a sale, despite the seller's efforts to construe it as a license.
And in any case, the agreement (whether a contract or a license) does not say anything about patents - specifically it does not say that the item is not licensed under Lexmark's patents. If the purchaser fails to abide by the contract terms, Lexmark's recourse is to sue the purchaser for breach of contract, not for patent infringement. At least, that was the legal situation before this Ninth Circuit ruling.