I agree, I generally think it's more sensible to code what you mean. If the purpose of a branch is to terminate the function early, why not just terminate the function with a return rather than introduce complexity within the function flow...
It just bugs me when people try to justify stylistic decisions like that with pseudo-technical arguments. Don't mean to single out your statement, it just happened to be the one that I responded to. I've seen them on both sides of almost every stylistic issue -- everyone's got some "once-upon-a-time" efficiency argument to support their favorite style... most of these "issues" are at best woefully out of date, but live on solely in coding standards.
The truth (IMO) is that code should be styled for humans to understand. O(1) efficiency sacrifices in favor of readability are usually a good trade-off until a profiler proves it's a bottleneck. The rub is defining "readability." (FWIW, I mostly support loose "common sense" style guidelines-- define whether indents are tabs or spaces and leave it to code reviews to catch egregious style violations)
Creating a return variable is also taking up more space in memory
That's probably not true. I'm not 100% sure for Java, but any decent C compiler takes the existence of a variable as a suggestion rather than a requirement (unless you declare it volatile, or under other special circumstances). It'll detect that you're just setting the variable and returning without checking it again, and probably generate very similar code whether you use split returns or not. AFAIK there's no strong general technical reason to prefer one return convention over the other, it's purely a style/debugging/human comprehension issue.
Good comment. IMO, a requirement (necessary, probably not sufficient but close) for creating a function out of a chunk of code is that its behavior be describable in English (or your human language of choice). That is, a function should do something meaningful in a higher level way. "Meaningless" logical bit-twiddling should be kept within functions.
I think it's ok if the function is only called once, as long as this criterion is met. This tends to cause things that are functions to be things that are reasonably likely to be used again; they're at least chunks that would show up in pseudo-code.
By the way, The Messiah had promised to filibuster the bill until it was dropped. It only takes one Senator to do that.
Actually, it takes 41 -- it only takes 60 votes to terminate a filibuster. This bill passed with 69 votes.
Maybe it's unlikely that some of the Dems who voted for the bill would have broken ranks with their party's candidate to vote to terminate his filibuster, but at the least it would have used up a sizable quantity of his political capital. While I'm extremely disappointed in Obama's (and the rest of Congress's) vote, it's not a simple decision.
While I'm extremely disappointed in his support for this bill, I don't know of any alternative candidate to vote for who did not also vote for it (where by alternative candidate I mean someone who has a real shot at winning).
While I agree with your point, I don't think it has particular bearing on what I'm saying. If their motion has factual basis in this case, then I think it should be granted, regardless of problems with their other cases. That has a giant "if" in it, though, and depends on facts of this particular case.
That said, I wouldn't argue against sanctions against them for an abusive pattern of lawsuits. That's a separate issue, though-- I don't think the right way to do this is to deny otherwise legitimate (if that's the case here) motions in some of their suits.
Anyway, I think we're talking about different things here.
My problem with KDE 4 was the size of the panel and the decorations. I'm on a laptop with limited screen real estate, so I like window manager decorations to be as small as possible. I wasn't able to reduce these without breaking all kinds of things.
The last thing I tried, might have been a beta of 4.1, don't recall, was closer. I think I could resize the panel to be smaller vertically, but this screwed up a lot of things visually. I'll wait, I'm happy with KDE 3.5.
If the prosecution can not prove a defendant guilty, then they should need to compensate them for the losses they incured as part of the suit, reimbursing them for legal fees and time.
That's a big issue and there are arguments both for and against it. Rather than open that debate here, I'll just note that the US is not a loser-pays system by default. Attorney's fees may be awarded to the defense, but as I understand it this is rare and is usually the result of malfeasance by the plaintiff. As you and others have noted, the RIAA does seem to be engaging in a campaign of questionable legality. However, I'm not really addressing that issue here.
I'm merely trying to point out that there are circumstances where the plaintiff may decide it's just not worth it to pursue the matter further. If this is simply because they're going to lose, then sure, they should dismiss with prejudice and consider awarding attorney's fees if it seems that their entire case was illegitimate. And, certainly, the case law as cited in the article seems to support this presumption. The RIAA, however, has argued that it's not simply that they're afraid they're going to lose, it's that they're afraid their going to lose due to abuse of the system by the defense. If (if if if!) that's true, then it seems unjust to require them to continue the case simply because they brought it. All lawsuits carry some risk of failure and if (if if if!) the defense was misbehaving, it will surely increase the cost of litigation through no fault of the plaintiff. It's quite reasonable that this may cause them to reconsider the cost/benefit scenario.
Anyway, I hope their position is without merit and they get slapped, but I don't think we can conclude this based solely on their motion and the defense's response. It depends on facts of the case, which I'm sure not up to date on.
A plaintiff that truly believes their case is fully with merit would follow this, unless they believe there is (now) going to be some harm to them to pursue it further (like actually losing, or worse).
I don't think you can draw that conclusion. For example, it's also possible that the stonewalling has served its purpose and the case is no longer worth their expense. Perhaps it's hard to imagine this for the RIAA, but consider a smaller-budgeted plaintiff suing a larger entity. If the defense's procedural games exhaust the resources of the plaintiff, they will have no choice but to drop the suit. An automatic award of defense costs would not seem fair.
After all, the RIAA is asking for sanctions against both the defendant and her counsel. I don't mean to speak to the merits, I'm just pointing out that they are making a request that's consistent with their rationale for dismissal without prejudice. They can (consistent with their motion as I understand it) claim that they've simply made a cost/benefit analysis and due to the defense's misbehavior (in their view), they are no longer interested in pursuing damages at this time.
Remember also that the US legal system does not automatically award legal costs to the loser of such a lawsuit. In other words, it takes more than simply failing to prove your case to entitle the defense to reimbursement.
All that said, I agree with your assessment of the RIAA's tactics and would love to see them smacked down. However, I'd like to be sure it's done in a completely honest, open way. If their claims are as meritless as the defense argues, I have no doubt that we'll see that.
You mean the right to remain silent? Or the right to an attorney? If you accuse me of something, the burden of proof is on you. I don't have to say a damn word.
No, I mean stonewalling by refusing to act in good faith during discovery. You have a legal obligation to comply with these requests. You don't have to *say* anything, but producing evidence that proves your guilt is not the same as incriminating yourself with respect to the 5th Amendment. If it's subpoenaed and it exists, you must provide it.
But, like I said, the RIAA made a claim that, on its face, is not without merit. Perhaps the facts of this case are such that it is a frivolous claim-- that's certainly what the article and common/. wisdom suggest. However, remember that the article and the various links are provided by the defense counsel, and/. is hardly free from anti-RIAA bias.
Regardless, my point is that a blanket statement that "The plaintiff must always see every case to completion," while perhaps satisfying in this case, will not always serve the interests of justice. For example, if the claims in the RIAA's motion are true as they're expressed there, then it is probably a reasonable request. As I said, I don't know enough about the details to try to do the job of the court, and that's not my point anyway.
It's tempting to make a blanket statement like that, especially when the plaintiff is perceived as the slimy evil that is the RIAA. However, it's difficult to make such a blanket declaration without creating injustices. In this case, if you read the RIAA's motion to dismiss without prejudice, which is linked from the article, they claim that Ms. Lindor and her relatives and counsel (aka NewYorkCountryLawyer, the submitter) have impeded/deceived them during discovery. Their claim is that as a result, they are unable to mount a case -- i.e., they're not just taking their ball and going home, they've simply decided that it's not going to be possible despite their (claim of) good faith effort to resolve the issues.
If this is true, then it's unreasonable to expect them to pay the costs of the defendant -- that would only encourage defense by stonewalling.
Thus, I don't think you can automatically say the plaintiff must see the case to conclusion to avoid funding the defense. As the case law seems to say (based on the links in the article), there's a presumption that they will be on the hook for costs, but there are certainly circumstances where justice is best served by allowing the case to be dropped without such an award.
And, hopefully obviously, I'm not up to date on the details of the case, and I'm not claiming that the RIAA is correct in its claims. Rather, it is simply those claims and the court needs to decide on their merit. The defense clearly disagrees with their assertions. It will be interesting to see how this turns out.
My old emacs color scheme was yellow on midnight blue. I'd read somewhere that yellow on blue was the highest effective color contrast, so tried it out and liked it. Unfortunately, this went to hell when syntax highlighting became the norm so now I just use the default black-on-white since I didn't want to redefine all the syntax colors to be readable on the blue background.
I still prefer light grey on black for terminal sessions.
To clarify what I think you're saying in a way that may be relevant to the grandparent post, the notebook is not a legal document by simple virtue of being a "scientist's notebook." Rather, a lab notebook can be a legally significant document in some legal proceedings. It is only of value in this context if you can demonstrate a consistent adherence to certain standards.
For example, if you are attempting to overturn someone else's patent due to your own prior work, your notebook describing and dating your ideas/tests may be of value. If you're able to produce a career's worth of similar notebooks describing your other work, that will lend credibility to the contents of your notebook.
You've reminded me why I stopped posting on slashdot a couple years ago, but you're pathetic enough that this is kind of amusing to me. (I know, I know, don't feed the trolls..) I'll try to put this in terms you can wrap your pathetic macaroni of a noodle around, ok? Stop me if I use a word that's too big for you.
Other people have written software. Some is available only under a proprietary license. Some is available under the GPL. Some uses other free licenses of various flavors or in the public domain. Since you
You, for whatever reason, would like to distribute software someone else has written. Maybe you have a hardware product in need of an OS, maybe you need some libraries to build your high-level software package, whatever. Let's consider your options under the various licensing schemes.
If the software you want to distribute is available under a proprietary license you can:
A) forget it and write it yourself B) try to negotiate with the owner and obtain a license for closed-source distribution C) distribute without a license and hope you don't get caught
Under B, you will probably end up paying for the license, if the owner is willing to grant one in the first place. Depending on the product, this may or may not be possible. Under C, if you are caught, you'll at least be financially liable unless you can work out terms that satisfy the owner (probably unlikely).
Now, suppose it's available under the GPL. You can:
A) forget it and write it yourself B) try to negotiate with the owner to obtain a license for closed-source distribution C) distribute without a license and hope you don't get caught D) accept the GPL and legally distribute it, but you must open your source
Notice that the GPL has *granted* you an extra option. Now, tell me what it's taken away.
You can only license proprietary software if the author/rights owner agrees to your price. Nothing prevents the owner/rights owner of GPL software from separately licensing it to you under a closed license if you can convince him to do so. The GPL takes nothing away.
That is, unless you're foolish enough not to understand who owns the software you're packaging with your own. As the grandparent posted, however, with the GPL you do have the *option* to open things up and probably avoid a financial settlement. If you can't afford that *and* you can't afford to settle with the rights owner, the GPL didn't do that to you. You did it to yourself.
Think about it this way. Do you think you'd be better off if you sold an embedded hardware project with unlicensed copies of QNX and this was discovered later? Would the proprietary license protect you from the costs of your blunder any more than the GPL would?
Your explanation is pretty lacking and your allusion to Nyquist (?) in your last paragraph is misleading at best. Describing Nyquist's theorem as compensation for the shortcomings of sampling is something I've never heard before... It really describes the requirement for capturing all the information in a finite bandwidth. Your ear is a physical transducer with finite frequency response, so you don't need infinite bandwidth anyway. All your reproduction has to be is better than the ear and you'll never notice the difference. If you don't think 20 kHz is sufficient bandwidth, then increase it to whatever you like. If you don't think 16 bits is sufficient, use 32. I'll not argue that the CD spec is the ultimate ideal in audio reproduction, but the point is that you can achieve any performance level through a sampled digital system. Plus, you can copy it without loss.
Note as well that your superior analog system has its own engineering limitations. The speakers and audio amplifiers are not perfectly linear, the vinyl flexes and warps with temperature and handling. Your needle and transducer has mass that will roll off the high end of your signal. You can only have so sharp a step in amplitude before the needle will jump. Et cetera.
Perhaps, but information is still lost.:-) Practically speaking, if you do any scaling or processing on that data, you'll notice the missing information, thus higher resolution ADCs for recording...
replying to myself... the reason the bass tends to get muddied has nothing to do with the sample rate which is easily high enough to perfectly reconstruct those tones from analog samples. It's the quantization that'll screw up the bass.
That's only true if you take an analog sample at each interval and use a perfect reconstruction filter. Once you restrict it to 16 bits, say, you're losing information.
To reproduce that sine wave after it has been digitized is a herculean task. Consider a tone that rings at 10khz, it has a periodicity.1 millisecond. To accurately digitize and then reproduce that sine wave perfectly, in 1/10th of a millisecond it needs to be sampled several thousand times. Remember, digital is either on or off, but a sine wave is an analog curve and to properly record that curve digitally it takes a huge number of data points. You're ignoring Nyquist's theorem and the reconstruction filter. If your signal is really bandlimited at 20 kHz and sampled at 44.1 kHz, the reconstruction filter can take the 4.41 points you get per 10 kHz cycle and perfectly reconstruct that original sine wave. There's no need to sample it thousands of times per period because there is no need to consider any signal that varies faster than 20 kHz. Those 4.41 samples per cycle exactly specify the output waveform in perfect mathematical precision.
There are a few reasons why CDs might sound worse than analog reproduction. Actually building that ideal reconstruction filter is hard. Generating at a perfectly uniform sampling frequency is kind of hard (though not nearly as hard as the first). Finally, audio isn't exactly perfectly bandlimited at a sharp 20 kHz cutoff, so it's possible that perceptual information is being rejected at the pre-sampling filter.
To sample the low end of the spectrum at a rate fast enough to accurately sample the high end of the spectrum would grossly over sample and this is why digital bass reproduction tends to be muddy. It's all sampled at the same rate...
I think what he's saying is that work ethic is a valuable thing. The guy who breezed through with A's may not have one. The guy who struggled to churn out C's has the work ethic. Assuming he's bright enough to learn from those experiences, his work ethic will be more valuable to the company. As is being discussed elsewhere in this thread at length, good grades don't instantly translate into solid job skills.
I'm sure he would prefer someone with a work ethic who could also breeze to solid A's. However, given a choice, he'd prefer reasonable skills and work ethic over an exceptionally-skilled slacker.
In my experience, this is not true of small companies, or at least less so than larger companies. If your whole workforce is 20 people, you can't tolerate swapping any of them out with any regularity. On the flip side, you are likely to need long hours occasionally since you can't spread unexpected critical tasks over as many people. Long hours are pretty hard to avoid for start-up to small companies and don't necessarily indicate exploitation.
A very very very small fraction of emails end up in a postmaster's box. That hardly invalidates the legal expectation of privacy. When an arrest is made on evidence found in a bounced email, post the story to slashdot... Grepping the mail spool? As the article points out, the courts specifically distinguished between sender/date (and possibly subject) meta-data and message content, so that's also irrelevant.
Do your job professionally. Just because you have access to something, doesn't give you any right to snoop. If you happen across clear evidence of illegal activities during the normal course of network maintenance, that's one thing. You'd be within your rights (and perhaps even required) to report that to the authorities. Engaging in a systematic filtering / collection of content is quite another, and that's what we're talking about here. And, specifically, by the government. If you, as a private citizen, decided to log and mine all the emails going through your server, that again would be a very different situation.
Beyond illegal, it's just plain BORING. Can you imagine how sad your life would be if you spiced it up by reading other people's frickin post cards?? I mean, it's hard to imagine being that miserable unless you're a postal wor... oh wait.
In all seriousness, don't suppose you have a link to back up the illegality of reading post cards, do you? I suspect you're right, but would just love to smack all the "but email is just a post card" schmucks with something concrete.
Slashdot Mirror
I agree, I generally think it's more sensible to code what you mean. If the purpose of a branch is to terminate the function early, why not just terminate the function with a return rather than introduce complexity within the function flow...
It just bugs me when people try to justify stylistic decisions like that with pseudo-technical arguments. Don't mean to single out your statement, it just happened to be the one that I responded to. I've seen them on both sides of almost every stylistic issue -- everyone's got some "once-upon-a-time" efficiency argument to support their favorite style... most of these "issues" are at best woefully out of date, but live on solely in coding standards.
The truth (IMO) is that code should be styled for humans to understand. O(1) efficiency sacrifices in favor of readability are usually a good trade-off until a profiler proves it's a bottleneck. The rub is defining "readability." (FWIW, I mostly support loose "common sense" style guidelines-- define whether indents are tabs or spaces and leave it to code reviews to catch egregious style violations)
Creating a return variable is also taking up more space in memory
That's probably not true. I'm not 100% sure for Java, but any decent C compiler takes the existence of a variable as a suggestion rather than a requirement (unless you declare it volatile, or under other special circumstances). It'll detect that you're just setting the variable and returning without checking it again, and probably generate very similar code whether you use split returns or not. AFAIK there's no strong general technical reason to prefer one return convention over the other, it's purely a style/debugging/human comprehension issue.
Good comment. IMO, a requirement (necessary, probably not sufficient but close) for creating a function out of a chunk of code is that its behavior be describable in English (or your human language of choice). That is, a function should do something meaningful in a higher level way. "Meaningless" logical bit-twiddling should be kept within functions.
I think it's ok if the function is only called once, as long as this criterion is met. This tends to cause things that are functions to be things that are reasonably likely to be used again; they're at least chunks that would show up in pseudo-code.
Fuck Steven Tyler with a chainsaw. What he produces, along with the idiots at Metallica is not music. BUT, they have found an audience.
Sorry, didn't realize we had to run things by you to figure out if it's music or not. Thanks for clarifying!
By the way, The Messiah had promised to filibuster the bill until it was dropped. It only takes one Senator to do that.
Actually, it takes 41 -- it only takes 60 votes to terminate a filibuster. This bill passed with 69 votes.
Maybe it's unlikely that some of the Dems who voted for the bill would have broken ranks with their party's candidate to vote to terminate his filibuster, but at the least it would have used up a sizable quantity of his political capital. While I'm extremely disappointed in Obama's (and the rest of Congress's) vote, it's not a simple decision.
Whom are you supporting instead?
While I'm extremely disappointed in his support for this bill, I don't know of any alternative candidate to vote for who did not also vote for it (where by alternative candidate I mean someone who has a real shot at winning).
While I agree with your point, I don't think it has particular bearing on what I'm saying. If their motion has factual basis in this case, then I think it should be granted, regardless of problems with their other cases. That has a giant "if" in it, though, and depends on facts of this particular case.
That said, I wouldn't argue against sanctions against them for an abusive pattern of lawsuits. That's a separate issue, though-- I don't think the right way to do this is to deny otherwise legitimate (if that's the case here) motions in some of their suits.
Anyway, I think we're talking about different things here.
My problem with KDE 4 was the size of the panel and the decorations. I'm on a laptop with limited screen real estate, so I like window manager decorations to be as small as possible. I wasn't able to reduce these without breaking all kinds of things.
The last thing I tried, might have been a beta of 4.1, don't recall, was closer. I think I could resize the panel to be smaller vertically, but this screwed up a lot of things visually. I'll wait, I'm happy with KDE 3.5.
If the prosecution can not prove a defendant guilty, then they should need to compensate them for the losses they incured as part of the suit, reimbursing them for legal fees and time.
That's a big issue and there are arguments both for and against it. Rather than open that debate here, I'll just note that the US is not a loser-pays system by default. Attorney's fees may be awarded to the defense, but as I understand it this is rare and is usually the result of malfeasance by the plaintiff. As you and others have noted, the RIAA does seem to be engaging in a campaign of questionable legality. However, I'm not really addressing that issue here.
I'm merely trying to point out that there are circumstances where the plaintiff may decide it's just not worth it to pursue the matter further. If this is simply because they're going to lose, then sure, they should dismiss with prejudice and consider awarding attorney's fees if it seems that their entire case was illegitimate. And, certainly, the case law as cited in the article seems to support this presumption. The RIAA, however, has argued that it's not simply that they're afraid they're going to lose, it's that they're afraid their going to lose due to abuse of the system by the defense. If (if if if!) that's true, then it seems unjust to require them to continue the case simply because they brought it. All lawsuits carry some risk of failure and if (if if if!) the defense was misbehaving, it will surely increase the cost of litigation through no fault of the plaintiff. It's quite reasonable that this may cause them to reconsider the cost/benefit scenario.
Anyway, I hope their position is without merit and they get slapped, but I don't think we can conclude this based solely on their motion and the defense's response. It depends on facts of the case, which I'm sure not up to date on.
A plaintiff that truly believes their case is fully with merit would follow this, unless they believe there is (now) going to be some harm to them to pursue it further (like actually losing, or worse).
I don't think you can draw that conclusion. For example, it's also possible that the stonewalling has served its purpose and the case is no longer worth their expense. Perhaps it's hard to imagine this for the RIAA, but consider a smaller-budgeted plaintiff suing a larger entity. If the defense's procedural games exhaust the resources of the plaintiff, they will have no choice but to drop the suit. An automatic award of defense costs would not seem fair.
After all, the RIAA is asking for sanctions against both the defendant and her counsel. I don't mean to speak to the merits, I'm just pointing out that they are making a request that's consistent with their rationale for dismissal without prejudice. They can (consistent with their motion as I understand it) claim that they've simply made a cost/benefit analysis and due to the defense's misbehavior (in their view), they are no longer interested in pursuing damages at this time.
Remember also that the US legal system does not automatically award legal costs to the loser of such a lawsuit. In other words, it takes more than simply failing to prove your case to entitle the defense to reimbursement.
All that said, I agree with your assessment of the RIAA's tactics and would love to see them smacked down. However, I'd like to be sure it's done in a completely honest, open way. If their claims are as meritless as the defense argues, I have no doubt that we'll see that.
You mean the right to remain silent? Or the right to an attorney? If you accuse me of something, the burden of proof is on you. I don't have to say a damn word.
No, I mean stonewalling by refusing to act in good faith during discovery. You have a legal obligation to comply with these requests. You don't have to *say* anything, but producing evidence that proves your guilt is not the same as incriminating yourself with respect to the 5th Amendment. If it's subpoenaed and it exists, you must provide it.
But, like I said, the RIAA made a claim that, on its face, is not without merit. Perhaps the facts of this case are such that it is a frivolous claim-- that's certainly what the article and common /. wisdom suggest. However, remember that the article and the various links are provided by the defense counsel, and /. is hardly free from anti-RIAA bias.
Regardless, my point is that a blanket statement that "The plaintiff must always see every case to completion," while perhaps satisfying in this case, will not always serve the interests of justice. For example, if the claims in the RIAA's motion are true as they're expressed there, then it is probably a reasonable request. As I said, I don't know enough about the details to try to do the job of the court, and that's not my point anyway.
It's tempting to make a blanket statement like that, especially when the plaintiff is perceived as the slimy evil that is the RIAA. However, it's difficult to make such a blanket declaration without creating injustices. In this case, if you read the RIAA's motion to dismiss without prejudice, which is linked from the article, they claim that Ms. Lindor and her relatives and counsel (aka NewYorkCountryLawyer, the submitter) have impeded/deceived them during discovery. Their claim is that as a result, they are unable to mount a case -- i.e., they're not just taking their ball and going home, they've simply decided that it's not going to be possible despite their (claim of) good faith effort to resolve the issues.
If this is true, then it's unreasonable to expect them to pay the costs of the defendant -- that would only encourage defense by stonewalling.
Thus, I don't think you can automatically say the plaintiff must see the case to conclusion to avoid funding the defense. As the case law seems to say (based on the links in the article), there's a presumption that they will be on the hook for costs, but there are certainly circumstances where justice is best served by allowing the case to be dropped without such an award.
And, hopefully obviously, I'm not up to date on the details of the case, and I'm not claiming that the RIAA is correct in its claims. Rather, it is simply those claims and the court needs to decide on their merit. The defense clearly disagrees with their assertions. It will be interesting to see how this turns out.
My old emacs color scheme was yellow on midnight blue. I'd read somewhere that yellow on blue was the highest effective color contrast, so tried it out and liked it. Unfortunately, this went to hell when syntax highlighting became the norm so now I just use the default black-on-white since I didn't want to redefine all the syntax colors to be readable on the blue background.
I still prefer light grey on black for terminal sessions.
To clarify what I think you're saying in a way that may be relevant to the grandparent post, the notebook is not a legal document by simple virtue of being a "scientist's notebook." Rather, a lab notebook can be a legally significant document in some legal proceedings. It is only of value in this context if you can demonstrate a consistent adherence to certain standards.
For example, if you are attempting to overturn someone else's patent due to your own prior work, your notebook describing and dating your ideas/tests may be of value. If you're able to produce a career's worth of similar notebooks describing your other work, that will lend credibility to the contents of your notebook.
You've reminded me why I stopped posting on slashdot a couple years ago, but you're pathetic enough that this is kind of amusing to me. (I know, I know, don't feed the trolls..) I'll try to put this in terms you can wrap your pathetic macaroni of a noodle around, ok? Stop me if I use a word that's too big for you.
Other people have written software. Some is available only under a proprietary license. Some is available under the GPL. Some uses other free licenses of various flavors or in the public domain. Since you
You, for whatever reason, would like to distribute software someone else has written. Maybe you have a hardware product in need of an OS, maybe you need some libraries to build your high-level software package, whatever. Let's consider your options under the various licensing schemes.
If the software you want to distribute is available under a proprietary license you can:
A) forget it and write it yourself
B) try to negotiate with the owner and obtain a license for closed-source distribution
C) distribute without a license and hope you don't get caught
Under B, you will probably end up paying for the license, if the owner is willing to grant one in the first place. Depending on the product, this may or may not be possible. Under C, if you are caught, you'll at least be financially liable unless you can work out terms that satisfy the owner (probably unlikely).
Now, suppose it's available under the GPL. You can:
A) forget it and write it yourself
B) try to negotiate with the owner to obtain a license for closed-source distribution
C) distribute without a license and hope you don't get caught
D) accept the GPL and legally distribute it, but you must open your source
Notice that the GPL has *granted* you an extra option. Now, tell me what it's taken away.
You can only license proprietary software if the author/rights owner agrees to your price. Nothing prevents the owner/rights owner of GPL software from separately licensing it to you under a closed license if you can convince him to do so. The GPL takes nothing away.
That is, unless you're foolish enough not to understand who owns the software you're packaging with your own. As the grandparent posted, however, with the GPL you do have the *option* to open things up and probably avoid a financial settlement. If you can't afford that *and* you can't afford to settle with the rights owner, the GPL didn't do that to you. You did it to yourself.
Think about it this way. Do you think you'd be better off if you sold an embedded hardware project with unlicensed copies of QNX and this was discovered later? Would the proprietary license protect you from the costs of your blunder any more than the GPL would?
Your explanation is pretty lacking and your allusion to Nyquist (?) in your last paragraph is misleading at best. Describing Nyquist's theorem as compensation for the shortcomings of sampling is something I've never heard before... It really describes the requirement for capturing all the information in a finite bandwidth. Your ear is a physical transducer with finite frequency response, so you don't need infinite bandwidth anyway. All your reproduction has to be is better than the ear and you'll never notice the difference. If you don't think 20 kHz is sufficient bandwidth, then increase it to whatever you like. If you don't think 16 bits is sufficient, use 32. I'll not argue that the CD spec is the ultimate ideal in audio reproduction, but the point is that you can achieve any performance level through a sampled digital system. Plus, you can copy it without loss.
Note as well that your superior analog system has its own engineering limitations. The speakers and audio amplifiers are not perfectly linear, the vinyl flexes and warps with temperature and handling. Your needle and transducer has mass that will roll off the high end of your signal. You can only have so sharp a step in amplitude before the needle will jump. Et cetera.
Perhaps, but information is still lost. :-) Practically speaking, if you do any scaling or processing on that data, you'll notice the missing information, thus higher resolution ADCs for recording...
replying to myself... the reason the bass tends to get muddied has nothing to do with the sample rate which is easily high enough to perfectly reconstruct those tones from analog samples. It's the quantization that'll screw up the bass.
That's only true if you take an analog sample at each interval and use a perfect reconstruction filter. Once you restrict it to 16 bits, say, you're losing information.
There are a few reasons why CDs might sound worse than analog reproduction. Actually building that ideal reconstruction filter is hard. Generating at a perfectly uniform sampling frequency is kind of hard (though not nearly as hard as the first). Finally, audio isn't exactly perfectly bandlimited at a sharp 20 kHz cutoff, so it's possible that perceptual information is being rejected at the pre-sampling filter. To sample the low end of the spectrum at a rate fast enough to accurately sample the high end of the spectrum would grossly over sample and this is why digital bass reproduction tends to be muddy. It's all sampled at the same rate...
I think what he's saying is that work ethic is a valuable thing. The guy who breezed through with A's may not have one. The guy who struggled to churn out C's has the work ethic. Assuming he's bright enough to learn from those experiences, his work ethic will be more valuable to the company. As is being discussed elsewhere in this thread at length, good grades don't instantly translate into solid job skills.
I'm sure he would prefer someone with a work ethic who could also breeze to solid A's. However, given a choice, he'd prefer reasonable skills and work ethic over an exceptionally-skilled slacker.
In my experience, this is not true of small companies, or at least less so than larger companies. If your whole workforce is 20 people, you can't tolerate swapping any of them out with any regularity. On the flip side, you are likely to need long hours occasionally since you can't spread unexpected critical tasks over as many people. Long hours are pretty hard to avoid for start-up to small companies and don't necessarily indicate exploitation.
A very very very small fraction of emails end up in a postmaster's box. That hardly invalidates the legal expectation of privacy. When an arrest is made on evidence found in a bounced email, post the story to slashdot... Grepping the mail spool? As the article points out, the courts specifically distinguished between sender/date (and possibly subject) meta-data and message content, so that's also irrelevant.
Do your job professionally. Just because you have access to something, doesn't give you any right to snoop. If you happen across clear evidence of illegal activities during the normal course of network maintenance, that's one thing. You'd be within your rights (and perhaps even required) to report that to the authorities. Engaging in a systematic filtering / collection of content is quite another, and that's what we're talking about here. And, specifically, by the government. If you, as a private citizen, decided to log and mine all the emails going through your server, that again would be a very different situation.
Beyond illegal, it's just plain BORING. Can you imagine how sad your life would be if you spiced it up by reading other people's frickin post cards?? I mean, it's hard to imagine being that miserable unless you're a postal wor... oh wait.
In all seriousness, don't suppose you have a link to back up the illegality of reading post cards, do you? I suspect you're right, but would just love to smack all the "but email is just a post card" schmucks with something concrete.