Would you care to detail the evidence proving: "Right now the crime rate is high because infractions have not been punished harsh enough from parents and law."
It seems a reasonable assertion, but I know of no evidence that it is actually correct. I could construct many other plausible statements, e.g.: 1) Right now the crime rate is high because those commiting the crimes aren't punished, and others are. 2) Right now the crime rate is high because violence and crime are glorified in the media. 3) Right now the crime rate is high because the government has destroyed it's moral authority. 4) Right now the crime rate is high because economic disparity between rich and poor has escalated enormously. 5)... Well, you get the idea. There are an indefinitely large number of possible explanations, and more than one may be true. But because there are so many, any particular one of them needs evidence that it is actually the, or a major, correct choice.
Fingerprints have NEVER honestly been used as evidence. To use it honestly one would need to calculate the probability that the portion of a fingerprint that was found at the scene was only present on one person. This would NEVER yield a certainty.
What is actually done is even less that that. They take the sample(s) and analyse them for a number of particular features, and then they look for someone with those particular features. An accurate comparison would require that after they select a suspect, that they then do a more complete comparison to ensure that the portions of the fingerprint that were not deemed "features" also match.
OTOH, fingerprints definitely increase the probability that the correct person will be apprehended than if you were to select at random out of all those persons matching the other known parts of the evidence. This is true even if you use fingerprints as the first selection criterion, provided that you then ensure that all of the remaining features of the evidence also match.
Unfortunately, there is little evidence that the evidence is combined in a method designed to maximize accuracy rather than in a way designed to maximize the chance of conviction. This means that evidence that only provides a filtering effect is presented as if it were a certain path to the true villian. (In such cases I am of the opinion that the true villian is the prosecuting attorney...with the judge who tolerates such abuse of justice as a close second. For some reason they don't seem to want me on juries, but I occasionally have managed to sneak on without lying.)
OTOH, certainty isn't something that is found outside of religion and mathematics. I feel that the certainty in mathematics is frequently overstated. And with religion...let's just say there isn't very much evidence, and it feels like certainty for the feeling of certainty.
I'd say the only thing wrong with your post is that you need to remove the word "typically", but that would be cynical. (OTOH, you did say "the court system", so I think that I'd be correct. Only some individuals within the system are concerned with justice.)
That's her address at work. It's probably answered by a secretary, and certainly goes through a lot of very heavy spam filtering. (I doubt that she would even notice anything as different.)
The thing is, she may well not HAVE any other e-mail address. If she did, she'd recognize spam from a sample.
errata: Re:first: prove the correctness of your...
on
Open Source Math
·
· Score: 1
change: (This is one of that all variables aren't global... to: (This is one of the reasons that all variables aren't global...
Re:first: prove the correctness of your software
on
Open Source Math
·
· Score: 1
Unfortunately, this also fails.
Very simple flaws have remained in well known math proofs for centuries. Yes, it's infrequent, but it HAS happened...so you never know it isn't happening again. And even if you were to assert that it has never happened, you would have the problem of proving that it didn't happen THIS time. It's an endless recursion.
Also, computer hardware is, itself, not inerrant...so you're not only trusting the software, you're also trusting the hardware.
Testing isn't foolproof...but neither is any other approach.
In very simple situations one can be rather certain that a program is correct, but as the program complexity grows, the certainty must decrease. (This is one of that all variables aren't global...which would yield much simpler compilers, but much more complex programs.)
People who use Spark Ada claim that they can prove the correctness of their programs, but the program specifications weren't written in Spark Ada...so there's no proof that the program as implemented matches the original specifications.
Now if you want likelihood rather than certainty, we're talking about something altogether more reasonable. But even here it gets quite difficult to calculate what the probability is that the program is correct (and what the error bars are for THAT calculation).
Everything real has error bars, either recognized or ignored. It makes things a lot simpler to just ignore them, but this doesn't mean that that's a valid approach (although, if you've got a time limit [and there always is one], it can be the best available choice).
Remember that it was just this year that a flaw was pointed out in the standard algorithm for a binary search. That algorithm had been out there since sometime around 1960...and it was just corrected in the standard implementations this year. Don't be certain that anything less scrutinized doesn't also have hidden flaws. (This was a problem that only appeared when the number of items in the list being sorted overflowed one half the maximum that the index could represent...so probably nobody ever encountered it, but it was there! As I remember it was "(a + b)/2" rather than "a + (a - b)/2" with a and b both signed non-negative indicies.)
Re:first: prove the correctness of your software
on
Open Source Math
·
· Score: 1
In many cases there are correctness tests that are much simpler than the original proof. In such cases, the steps taken to derive the proof are not necessarily open. (OTOH, I'd hesitate to call such a proof a constructive proof rather than an existence proof.)
Neither open nor closed software can be known with any reliability to be without errors. You are saying that there is a sizeable program that doesn't have any bugs! If you fix a bug, then you run into the problem of KNOWING that you haven't introduced a new error, so you need to start checking all over.
As a method for dealing with this, large programs usually have "test suites". When the program passes all of the tests, it's deemed "good enough". As bugs are fixed, more pieces are added to the test suite, so more cases are handled correctly. I don't think, however, that one can ever reach the point where a sizeable program is known to be bug free...and if you were to reach that point, you'd need to move to a new processor/compiler combination, at which point you must start your testing again.
Probably mathematics beyond a certain level of complexity cannot be know with certainty to be correct, and beyond some slightly higher level it can be known to contain unexpected bugs. (This doesn't help with locating them!)
Re:Not necessarily bad in all cases...
on
Open Source Math
·
· Score: 1
I've encountered too many programs where the source code doesn't match the documentation. For some of your arguments, that's not fatal. If the entire proofs are made explicit, then you can argue that it's like not being able to peer inside the skull of the mathematician. In the cases, however, where you are depending on the results of computational steps (as in the four color proof), those steps need to be made open and explicit.
Pseudocode is not sufficient. You don't know that it actually reflects the code that was used.
That sounds like the explanation they gave at the time.
I read the original license, read their changes, and wasn't satisfied. But I didn't need to use that license, so I didn't ask a lawyer.
As for your question "Can you dynamically link to a GPL library with non-GPL code?", I try to avoid getting NEAR to the edges of legal agreements. If I had to do what you are suggesting, I'd either contact the licensor, and get their explicit approval, or get a lawyer and get HIS explict approval. So far I've never needed to investigate that disagreement (and I'm aware that many non-lawyers are in disagreement over this particular detail...so I count that as a part of what I know about the license).
This is technically almost correct...but only because the GPL was there before the CDDL was. Were the time order reversed you could make the exact same argument about the CDDL.
OTOH, the CDDL, as I remember it (not certain...I haven't kept up with Sun's license versions) had special advantages for the company that originated the software (or perhaps just for Sun). But this might have been an earlier license, or an earlier version of the CDDL. (The one I'm remembering was based off the MPL. I looked it over and decided that: a) It's not *THAT* bad. If I *must* use it I can, but I'm sure going to try to avoid it. b) I can avoid it, so I will.
For me one of the real advantages of the GPL is that I know what the license means without having to buy a lawyer. Buying a lawyer is a large overhead expense. If you already have one on staff, I can see why you might have different accounting rules...you don't have the same expense structure. So I have generally preferred to only use OSI approved licenses. With their recent changes, however, I guess I'll have to switch to only using licenses that they approved before 2005 (a bit of slack for safety and ease of memory) or which have been approved by the FSF. I'm not doctrinaire the way the FSF is, but so far they've been a trustworthy, if a bit over restrictive, guide.
OTOH, I'm trying to remember back to that old Sun license based around the MPL. They deleted a paragraph that seemed pretty important, and since I decided I could avoid the license, I don't remember exactly what it said, but it was about patents. All-in-all I think I trust the GPL (both versions 2 & 3) more than I trust the CDDL. IANAL, and perhaps that's why, but it's also the case that I read the licenses, and *THAT* may be why.
You are very fortunate that the people who put together Ubuntu gave more thought to it than you have. And doubly fortunate that the people who put together the Debian version that your version of Ubuntu gave it even more thought. And triply fortunate that...
I think it goes on for about eight levels. Possibly more. There ARE tradeoffs between ease of use and security. I'm not sure that opting for sudo was the correct choice for Ubuntu to make...but I'm not sure that it wasn't. To me it feels less secure, but that may be an illusion. Linux finally decided that automount was worth the risk, and that's usually correct. But that was considered for most of a decade (how seriously I don't know).
I understand, however, that MSWind has finally started to give security a bit of thought. Originally MSWind was reported as having as little security as possible. They actively removed it. (My understanding was that this was an effort to speed up the system.) They didn't even start giving lip-service to implementing security in MSWind until NT...and there wasn't noticable security in the customer line until XT(XP?), by which time I'd stopped using it for years. People tell me that currently MSWind has decent security. I don't know whether they are correct or not, as I've stopped using it. My inclination is to scoff, but this isn't based on any evidence, and may be incorrect. If it is secure, I doubt that it's because the customers want it...MS probably has some ulterior motive. (They've rarely done things to benefit the end user, though sometimes the end-user has benefited from things that they've done for other reasons [in which I include "to keep their market share from eroding"]).
I had heard that the authorizing bill expired. Perhaps your version of the story is more correct. In any case, between the time that your coverage expires and the time you reregister your number could be freely shared or sold, say, with people outside the country who wouldn't be bound by the reregistration.
Google has not repeatedly proven itself untrustwory. The US govt. has.
Therefore Google is more trusted than the US govt.
Do I trust Google without reservations? No. I wouldn't use Google for my data storage center. I don't put my e-mail on Google (or Yahoo, or Hotmail, or...).
This isn't quite privacy, but it's related. I don't trust remotely changable information to not be remotely changed to suite political or economic convenience. And I don't trust what they say about me to be honest, either, but I can't do much about that. Telling the truth never has much effect when you're speaking to someone with an agenda.
Some amount of risk acceptance is necessary. Risks don't go away just because you don't pay attention to them. But that's no reason to feed the monster.
This "privacy advice" is coming from someone speaking for a particularly untrustworthy government, but I wouldn't have trusted Clinton with my information either, and I won't trust the next one either. The government has an agenda, and it's only peripherally related to the purposes of the citizenry. This is true of all branches, but it's most true of the executive, and least true of the judicial. Something about running for office seems to instill corruption into the heart of political candidates. (This was always true, but the FCC decision that stations didn't need to provide equal access to electoral candidates really cranked up the level of corruption.)
It requires trusting the people who will be collecting the information. Experience proves that they are *NOT* trustworthy, and don't have your best interests at heart.
Even if you can't get total privacy, get what you can, and don't give up easily. Those who are trying to replace privacy with trusting large organizations are doing so because large organizations can be threatened by larger or more powerful (or even just more committed) organizations.
P.S.: Remember that "Do Not Call" list? That one shares your phone number with all telemarketers, so they'll know who not to call. It expires next year, and they've got your number.
I'm not sure those numbers mean what you say they mean, but *I've* stopped considering Novell products. (OTOH, there is no evidence that the people signing that petition were ever using Novell. Some of them probably were, but then a lot of people who didn't sign the petition have also moved away from Novell.)
I do hope that Eben Mogen's hopes are justified. But my choices aren't dependent on what a court may decide, or such. Novell is no longer considered for purchase.
Before the deal I had SUSE, Debian, and Ubuntu installed. Now I have Debian and Ubuntu. (Both before and after Debian was the main system. SUSE was there largely because I wanted to see what the noise was about, and partially to support Novell against SCOX.)
I don't think that's sarcasm. They are all (or almost all) guilty, though unindicted.
OTOH, you clearly realize that they AREN'T going to be locked up. That doesn't sound quite like sarcasm. It's also not exactly ironic (though it certainly has ironic overtones). Caustic or acerbit humor seems more accurate, but I can't think of a single word that expresses it.
In that case it's *really* strange that the version I installed on Debian was labelled compiled with gcj. And it's strange that the version on the Mac crawled, while The Gimp ran fine. Still, too many variables.
This is a part of the confusion over the term "open". "Free" would have been more appropriate, but would have had it's own sorts of confusion.
Sun's license was "open" in the sense that you could see the code. It wasn't "free" because you couldn't redistribute it. This made it unuseable by Linux distros. To avoid confusion over price, they said the code wasn't open. Technically the wrong term, but free would have caused as much, or more, confusion, and many more people would have thought that they were lying.
Yeah, I dumped Java because of this (and other reasons), but it wasn't out of misunderstanding. It was because it kept getting broken. (I also really detest the way that they handle files...even though I understand some of the reasons for their choices.)
P.S.: When I installed Sun Java on Linux, I could always get it to work except in places where others were also having trouble. gcj I have trouble with because of scanty documentation, and because I'm not really familiar with the linker, building library files, etc. I've usually been able to work through any adequately documented example, I just haven't been able to adapt it and keep it working. I think the gcj documenters expect one to be familiar with compiling with gcc and g++, which I'm not. (The documentation REALLY needs to be improved! It's almost absent.)
Sorry, but NetBeans has convinced me that even Sun selected specialist experts can't write a responsive GUI in Java. The reason why may be debateable, the fact is observable.
Also, I've heard lots of complaints about the responsiveness of OpenOffice. I'm presuming that these are from people who are using versions that aren't compiled with gcj...but if that's what makes the difference, then it's Java rather than Swing that's the problem. (N.B.: This *IS* a presumption. I've noticed that OpenOffice on the Mac is unresponsive, but it's running via X layered on top of the Mac OS. NeoOffice seems to be... sort of OK, but not comparable to a gcj compiled OpenOffice on Linux. Too many variables, but the position of the jvm is suspiciously central.
Well, the last time I checked you didn't need a driver's license in North Dakota. I think that was the only state that didn't require it.
(IIRC, if you don't require drivers licenses, the Feds won't put in any Interstate Highways...but North Dakota wouldn't get one anyway...so they didn't bother. If you didn't cap your speed limit, the Feds wouldn't but in any Interstate Highways...but...)
OTOH, do remember that North Dakota is a largely rural state with a low population. Things that make sense in urban settings don't make any sense there. One size does NOT fit all.
The first problem with that solution is that some people seem to be compelled to vandalize things if nobody's watching. A camera doesn't seem to count. (Well, actually the camera *IS* the first thing vandalized, so I suppose it DOES count, but not in a useful manner.)
Some societies handle this better than others, but I believe that all have a problem with this. (The US appears to me to average worse than Japan...but I never worked for a transportation company in Japan.) The amount of graffiti is probably a reasonable estimator...but vehicles travel, so something that gets defaced in one part of town will end up in another...which means that you can't have vandals ANYWHERE.
I can't imagine that they would react any other way, unless it was by just ignoring the accusations. They certainly couldn't admit it.
OTOH, I'm not sure that "tortous interference with a business relationship" is an international crime, so maybe they *could* have safely admitted it. But if they did I'd imagine it might show up in various legal hearing on illegal monopolies (illegal use of monopoly?).
MS: Campaigning for the role of most loathed international criminal conspiracy.
It's got prototypes, too, though you need to be a bit clever about it. (As usual, using prototypes adds overhead. Sigh!)
Arguing that Javascript is a complete language is rather like arguing that a Turing machine is a complete programming environment. Technically, it's a correct statement. Practically... well, it's a bit slow and clumsy.
N.B.: Python is slow compared to C, but it's NOT clumsy. Javascript is both slow and clumsy compared to Python. In an environment where the critical path component is the reactions of the end-user the "Slow" usually isn't significant. In an environment where you WANT to constrict what can be done, the clumsy can be a genuine benefit. (Just TRY to write a random access file manipulation routine in Javascript! Now imagine trusting such a file coming in from a randomly chosen web site...and you'll see the advantages of that being difficult. [I won't say that it's impossible, as I don't believe that it is...at least not if you assume the presence of a few standard libraries. Which is a pity. Java attempted to MAKE such a thing happening impossible. Perhaps it succeeded, as I haven't heard any dire warnings about allowing remote execution of Java applets recently. But Java's a relatively heavyweight language.])
Not sure about now, but awhile ago Apple was an excellent choice for a server machine...if you didn't worry about your budget. If you did, then either Linux or BSD were good choices: they could be at least as secure, and a lot more cost effective.
OTOH, if all you have are graphic designers...including a few that dabble in computer administration, then Apple is the BEST choice. Easier for novices to configure safely, and not THAT much more expensive. (Much less than an additional employee.)
However, be sure that you read the EULA. Apple has added that same "we have the right to add, delete, modify, or copy any file on your computer" that MS has, and I find that totally unacceptable. (This was in the EULAs to a quicktime upgrade pack...I don't know whether it's in the new basic EULAs, as I have stopped buying Apple.)
Slashdot Mirror
Would you care to detail the evidence proving:
...
"Right now the crime rate is high because infractions have not been punished harsh enough from parents and law."
It seems a reasonable assertion, but I know of no evidence that it is actually correct. I could construct many other plausible statements, e.g.:
1) Right now the crime rate is high because those commiting the crimes aren't punished, and others are.
2) Right now the crime rate is high because violence and crime are glorified in the media.
3) Right now the crime rate is high because the government has destroyed it's moral authority.
4) Right now the crime rate is high because economic disparity between rich and poor has escalated enormously.
5)
Well, you get the idea. There are an indefinitely large number of possible explanations, and more than one may be true. But because there are so many, any particular one of them needs evidence that it is actually the, or a major, correct choice.
Fingerprints have NEVER honestly been used as evidence. To use it honestly one would need to calculate the probability that the portion of a fingerprint that was found at the scene was only present on one person. This would NEVER yield a certainty.
What is actually done is even less that that. They take the sample(s) and analyse them for a number of particular features, and then they look for someone with those particular features. An accurate comparison would require that after they select a suspect, that they then do a more complete comparison to ensure that the portions of the fingerprint that were not deemed "features" also match.
OTOH, fingerprints definitely increase the probability that the correct person will be apprehended than if you were to select at random out of all those persons matching the other known parts of the evidence. This is true even if you use fingerprints as the first selection criterion, provided that you then ensure that all of the remaining features of the evidence also match.
Unfortunately, there is little evidence that the evidence is combined in a method designed to maximize accuracy rather than in a way designed to maximize the chance of conviction. This means that evidence that only provides a filtering effect is presented as if it were a certain path to the true villian. (In such cases I am of the opinion that the true villian is the prosecuting attorney...with the judge who tolerates such abuse of justice as a close second. For some reason they don't seem to want me on juries, but I occasionally have managed to sneak on without lying.)
OTOH, certainty isn't something that is found outside of religion and mathematics. I feel that the certainty in mathematics is frequently overstated. And with religion...let's just say there isn't very much evidence, and it feels like certainty for the feeling of certainty.
I'd say the only thing wrong with your post is that you need to remove the word "typically", but that would be cynical. (OTOH, you did say "the court system", so I think that I'd be correct. Only some individuals within the system are concerned with justice.)
That's her address at work. It's probably answered by a secretary, and certainly goes through a lot of very heavy spam filtering. (I doubt that she would even notice anything as different.)
The thing is, she may well not HAVE any other e-mail address. If she did, she'd recognize spam from a sample.
change:
(This is one of that all variables aren't global...
to:
(This is one of the reasons that all variables aren't global...
Unfortunately, this also fails.
Very simple flaws have remained in well known math proofs for centuries. Yes, it's infrequent, but it HAS happened...so you never know it isn't happening again. And even if you were to assert that it has never happened, you would have the problem of proving that it didn't happen THIS time. It's an endless recursion.
Also, computer hardware is, itself, not inerrant...so you're not only trusting the software, you're also trusting the hardware.
Testing isn't foolproof...but neither is any other approach.
In very simple situations one can be rather certain that a program is correct, but as the program complexity grows, the certainty must decrease. (This is one of that all variables aren't global...which would yield much simpler compilers, but much more complex programs.)
People who use Spark Ada claim that they can prove the correctness of their programs, but the program specifications weren't written in Spark Ada...so there's no proof that the program as implemented matches the original specifications.
Now if you want likelihood rather than certainty, we're talking about something altogether more reasonable. But even here it gets quite difficult to calculate what the probability is that the program is correct (and what the error bars are for THAT calculation).
Everything real has error bars, either recognized or ignored. It makes things a lot simpler to just ignore them, but this doesn't mean that that's a valid approach (although, if you've got a time limit [and there always is one], it can be the best available choice).
Remember that it was just this year that a flaw was pointed out in the standard algorithm for a binary search. That algorithm had been out there since sometime around 1960...and it was just corrected in the standard implementations this year. Don't be certain that anything less scrutinized doesn't also have hidden flaws. (This was a problem that only appeared when the number of items in the list being sorted overflowed one half the maximum that the index could represent...so probably nobody ever encountered it, but it was there! As I remember it was "(a + b)/2" rather than "a + (a - b)/2" with a and b both signed non-negative indicies.)
In many cases there are correctness tests that are much simpler than the original proof. In such cases, the steps taken to derive the proof are not necessarily open. (OTOH, I'd hesitate to call such a proof a constructive proof rather than an existence proof.)
Neither open nor closed software can be known with any reliability to be without errors. You are saying that there is a sizeable program that doesn't have any bugs! If you fix a bug, then you run into the problem of KNOWING that you haven't introduced a new error, so you need to start checking all over.
As a method for dealing with this, large programs usually have "test suites". When the program passes all of the tests, it's deemed "good enough". As bugs are fixed, more pieces are added to the test suite, so more cases are handled correctly. I don't think, however, that one can ever reach the point where a sizeable program is known to be bug free...and if you were to reach that point, you'd need to move to a new processor/compiler combination, at which point you must start your testing again.
Probably mathematics beyond a certain level of complexity cannot be know with certainty to be correct, and beyond some slightly higher level it can be known to contain unexpected bugs. (This doesn't help with locating them!)
I've encountered too many programs where the source code doesn't match the documentation. For some of your arguments, that's not fatal. If the entire proofs are made explicit, then you can argue that it's like not being able to peer inside the skull of the mathematician. In the cases, however, where you are depending on the results of computational steps (as in the four color proof), those steps need to be made open and explicit.
Pseudocode is not sufficient. You don't know that it actually reflects the code that was used.
That sounds like the explanation they gave at the time.
I read the original license, read their changes, and wasn't satisfied. But I didn't need to use that license, so I didn't ask a lawyer.
As for your question "Can you dynamically link to a GPL library with non-GPL code?", I try to avoid getting NEAR to the edges of legal agreements. If I had to do what you are suggesting, I'd either contact the licensor, and get their explicit approval, or get a lawyer and get HIS explict approval. So far I've never needed to investigate that disagreement (and I'm aware that many non-lawyers are in disagreement over this particular detail...so I count that as a part of what I know about the license).
This is technically almost correct...but only because the GPL was there before the CDDL was. Were the time order reversed you could make the exact same argument about the CDDL.
OTOH, the CDDL, as I remember it (not certain...I haven't kept up with Sun's license versions) had special advantages for the company that originated the software (or perhaps just for Sun). But this might have been an earlier license, or an earlier version of the CDDL. (The one I'm remembering was based off the MPL. I looked it over and decided that:
a) It's not *THAT* bad. If I *must* use it I can, but I'm sure going to try to avoid it.
b) I can avoid it, so I will.
For me one of the real advantages of the GPL is that I know what the license means without having to buy a lawyer. Buying a lawyer is a large overhead expense. If you already have one on staff, I can see why you might have different accounting rules...you don't have the same expense structure. So I have generally preferred to only use OSI approved licenses. With their recent changes, however, I guess I'll have to switch to only using licenses that they approved before 2005 (a bit of slack for safety and ease of memory) or which have been approved by the FSF. I'm not doctrinaire the way the FSF is, but so far they've been a trustworthy, if a bit over restrictive, guide.
OTOH, I'm trying to remember back to that old Sun license based around the MPL. They deleted a paragraph that seemed pretty important, and since I decided I could avoid the license, I don't remember exactly what it said, but it was about patents. All-in-all I think I trust the GPL (both versions 2 & 3) more than I trust the CDDL. IANAL, and perhaps that's why, but it's also the case that I read the licenses, and *THAT* may be why.
You are very fortunate that the people who put together Ubuntu gave more thought to it than you have.
And doubly fortunate that the people who put together the Debian version that your version of Ubuntu gave it even more thought.
And triply fortunate that...
I think it goes on for about eight levels. Possibly more. There ARE tradeoffs between ease of use and security. I'm not sure that opting for sudo was the correct choice for Ubuntu to make...but I'm not sure that it wasn't. To me it feels less secure, but that may be an illusion. Linux finally decided that automount was worth the risk, and that's usually correct. But that was considered for most of a decade (how seriously I don't know).
I understand, however, that MSWind has finally started to give security a bit of thought. Originally MSWind was reported as having as little security as possible. They actively removed it. (My understanding was that this was an effort to speed up the system.) They didn't even start giving lip-service to implementing security in MSWind until NT...and there wasn't noticable security in the customer line until XT(XP?), by which time I'd stopped using it for years. People tell me that currently MSWind has decent security. I don't know whether they are correct or not, as I've stopped using it. My inclination is to scoff, but this isn't based on any evidence, and may be incorrect. If it is secure, I doubt that it's because the customers want it...MS probably has some ulterior motive. (They've rarely done things to benefit the end user, though sometimes the end-user has benefited from things that they've done for other reasons [in which I include "to keep their market share from eroding"]).
I had heard that the authorizing bill expired. Perhaps your version of the story is more correct. In any case, between the time that your coverage expires and the time you reregister your number could be freely shared or sold, say, with people outside the country who wouldn't be bound by the reregistration.
Google has not repeatedly proven itself untrustwory.
The US govt. has.
Therefore Google is more trusted than the US govt.
Do I trust Google without reservations? No. I wouldn't use Google for my data storage center. I don't put my e-mail on Google (or Yahoo, or Hotmail, or...).
This isn't quite privacy, but it's related. I don't trust remotely changable information to not be remotely changed to suite political or economic convenience. And I don't trust what they say about me to be honest, either, but I can't do much about that. Telling the truth never has much effect when you're speaking to someone with an agenda.
Some amount of risk acceptance is necessary. Risks don't go away just because you don't pay attention to them. But that's no reason to feed the monster.
This "privacy advice" is coming from someone speaking for a particularly untrustworthy government, but I wouldn't have trusted Clinton with my information either, and I won't trust the next one either. The government has an agenda, and it's only peripherally related to the purposes of the citizenry. This is true of all branches, but it's most true of the executive, and least true of the judicial. Something about running for office seems to instill corruption into the heart of political candidates. (This was always true, but the FCC decision that stations didn't need to provide equal access to electoral candidates really cranked up the level of corruption.)
It requires trusting the people who will be collecting the information. Experience proves that they are *NOT* trustworthy, and don't have your best interests at heart.
Even if you can't get total privacy, get what you can, and don't give up easily. Those who are trying to replace privacy with trusting large organizations are doing so because large organizations can be threatened by larger or more powerful (or even just more committed) organizations.
P.S.: Remember that "Do Not Call" list? That one shares your phone number with all telemarketers, so they'll know who not to call. It expires next year, and they've got your number.
I'm not sure those numbers mean what you say they mean, but *I've* stopped considering Novell products. (OTOH, there is no evidence that the people signing that petition were ever using Novell. Some of them probably were, but then a lot of people who didn't sign the petition have also moved away from Novell.)
I do hope that Eben Mogen's hopes are justified. But my choices aren't dependent on what a court may decide, or such. Novell is no longer considered for purchase.
Before the deal I had SUSE, Debian, and Ubuntu installed. Now I have Debian and Ubuntu. (Both before and after Debian was the main system. SUSE was there largely because I wanted to see what the noise was about, and partially to support Novell against SCOX.)
I don't think that's sarcasm. They are all (or almost all) guilty, though unindicted.
OTOH, you clearly realize that they AREN'T going to be locked up. That doesn't sound quite like sarcasm. It's also not exactly ironic (though it certainly has ironic overtones). Caustic or acerbit humor seems more accurate, but I can't think of a single word that expresses it.
GHB *is* GBH!
In that case it's *really* strange that the version I installed on Debian was labelled compiled with gcj. And it's strange that the version on the Mac crawled, while The Gimp ran fine. Still, too many variables.
This is a part of the confusion over the term "open". "Free" would have been more appropriate, but would have had it's own sorts of confusion.
Sun's license was "open" in the sense that you could see the code. It wasn't "free" because you couldn't redistribute it. This made it unuseable by Linux distros. To avoid confusion over price, they said the code wasn't open. Technically the wrong term, but free would have caused as much, or more, confusion, and many more people would have thought that they were lying.
Yeah, I dumped Java because of this (and other reasons), but it wasn't out of misunderstanding. It was because it kept getting broken. (I also really detest the way that they handle files...even though I understand some of the reasons for their choices.)
P.S.: When I installed Sun Java on Linux, I could always get it to work except in places where others were also having trouble. gcj I have trouble with because of scanty documentation, and because I'm not really familiar with the linker, building library files, etc. I've usually been able to work through any adequately documented example, I just haven't been able to adapt it and keep it working. I think the gcj documenters expect one to be familiar with compiling with gcc and g++, which I'm not. (The documentation REALLY needs to be improved! It's almost absent.)
Like NetBeans?
... sort of OK, but not comparable to a gcj compiled OpenOffice on Linux. Too many variables, but the position of the jvm is suspiciously central.
Sorry, but NetBeans has convinced me that even Sun selected specialist experts can't write a responsive GUI in Java. The reason why may be debateable, the fact is observable.
Also, I've heard lots of complaints about the responsiveness of OpenOffice. I'm presuming that these are from people who are using versions that aren't compiled with gcj...but if that's what makes the difference, then it's Java rather than Swing that's the problem. (N.B.: This *IS* a presumption. I've noticed that OpenOffice on the Mac is unresponsive, but it's running via X layered on top of the Mac OS. NeoOffice seems to be
Well, the last time I checked you didn't need a driver's license in North Dakota. I think that was the only state that didn't require it.
(IIRC, if you don't require drivers licenses, the Feds won't put in any Interstate Highways...but North Dakota wouldn't get one anyway...so they didn't bother. If you didn't cap your speed limit, the Feds wouldn't but in any Interstate Highways...but...)
OTOH, do remember that North Dakota is a largely rural state with a low population. Things that make sense in urban settings don't make any sense there. One size does NOT fit all.
The first problem with that solution is that some people seem to be compelled to vandalize things if nobody's watching. A camera doesn't seem to count. (Well, actually the camera *IS* the first thing vandalized, so I suppose it DOES count, but not in a useful manner.)
Some societies handle this better than others, but I believe that all have a problem with this. (The US appears to me to average worse than Japan...but I never worked for a transportation company in Japan.) The amount of graffiti is probably a reasonable estimator...but vehicles travel, so something that gets defaced in one part of town will end up in another...which means that you can't have vandals ANYWHERE.
I can't imagine that they would react any other way, unless it was by just ignoring the accusations. They certainly couldn't admit it.
OTOH, I'm not sure that "tortous interference with a business relationship" is an international crime, so maybe they *could* have safely admitted it. But if they did I'd imagine it might show up in various legal hearing on illegal monopolies (illegal use of monopoly?).
MS: Campaigning for the role of most loathed international criminal conspiracy.
It's got prototypes, too, though you need to be a bit clever about it. (As usual, using prototypes adds overhead. Sigh!)
Arguing that Javascript is a complete language is rather like arguing that a Turing machine is a complete programming environment. Technically, it's a correct statement. Practically... well, it's a bit slow and clumsy.
N.B.: Python is slow compared to C, but it's NOT clumsy. Javascript is both slow and clumsy compared to Python. In an environment where the critical path component is the reactions of the end-user the "Slow" usually isn't significant. In an environment where you WANT to constrict what can be done, the clumsy can be a genuine benefit. (Just TRY to write a random access file manipulation routine in Javascript! Now imagine trusting such a file coming in from a randomly chosen web site...and you'll see the advantages of that being difficult. [I won't say that it's impossible, as I don't believe that it is...at least not if you assume the presence of a few standard libraries. Which is a pity. Java attempted to MAKE such a thing happening impossible. Perhaps it succeeded, as I haven't heard any dire warnings about allowing remote execution of Java applets recently. But Java's a relatively heavyweight language.])
Not sure about now, but awhile ago Apple was an excellent choice for a server machine...if you didn't worry about your budget. If you did, then either Linux or BSD were good choices: they could be at least as secure, and a lot more cost effective.
OTOH, if all you have are graphic designers...including a few that dabble in computer administration, then Apple is the BEST choice. Easier for novices to configure safely, and not THAT much more expensive. (Much less than an additional employee.)
However, be sure that you read the EULA. Apple has added that same "we have the right to add, delete, modify, or copy any file on your computer" that MS has, and I find that totally unacceptable. (This was in the EULAs to a quicktime upgrade pack...I don't know whether it's in the new basic EULAs, as I have stopped buying Apple.)