This is the first bit of sense regarding ths matter I've seen.
Those that tease, taunt. and torture, should expect that some of the people they treat this way will react by killing others.
If you constantly beat a dog with a stick, the next time he sees a child with a similar strick, he will likely attack her.
Does this mean that the killers are blameless? No, but those that do things that are "wrong, but silently accepted" should be held accountable as well for encouraging others to become unstable.
This isn't the same as admonishing people to "walk on eggshells". Just because something one does may annoy somebody doesn't make it wrong. However, if one would not wish to be on the receiving end of what they dish out, they should be held accountable for the consequences.
Original post: the major players here are someone who didn't bother or didn't succeed to instill a sense of morality
My response: Er, no. The major players here are those whose teasing and taunts drove someone to lash back.
Your analysis: All right, you're saying that the real perpetrators of this crime are the various Jocks, Soshs, and Primadonnas of the Colombine HS social structure because they picked on some kids they thought were weird.
No. I used the term major players, to refer to those who could prevent such a tragedy. I disagree with the person who's post I followed up about just who the major players are. If I meant to write about the alleged (remember, even they are innocent until proven guilty) perpetrators of the crime, I would have used different words.
Since your understanding of what I wrote is in error, your further analysis is invalid.
the major players here are someone who didn't bother or didn't succeed to instill a sense of morality
Er, no. The major players here are those whose teasing and taunts drove someone to lash back.
What sense is there to instill a sense of morality and decency in someone if OTHERS will not act accordingly?
"Do unto others as you would have them do unto you" becomes perverted into "Do unto others as they have done unto you".
When some people disregard other's freedom to enjoy so-called alternate lifestyles, they should not be surprised when their freedoms are totally disregarded in return.
The overblown response of murder to teasing is perfectly in-line with the reasoning that the best response is a strong one (and why the U.S. floundered in Vietnam and won in the Gulf).
Is such a reaction right or justified? Certainly not, but the risk of it happening is to be expected. It is truely unfortunate that the response is directed so poorly at the instigators, but then again, who among is us innocent when we allow harrassment to persist?
The best premptive measures are those that serve to nip harrassment in the bud, and thus not giving rise to any form of retaliation, overblown or otherwise.
While I do not advocate taking the law into one's own hands to settle a dispute, neither do I care much when the rights of those who are disrespectful of others are themselves trampled upon. I do wish, however, that when such an overblown response occurs, it occur with greater regard for who the recipient is. Sadly, I do not ever expect this to be the case.
Its not that simple: layering KDE on top of a Debian distribution requires identifying HOW to effect such layering: where to put stuff, etc.
While KDE is not free, the information that makes the layering seamless can be.
Corel's distribution necessarily must be more than just Debian GNU/Linux + KDE. If not, then you could just get Debian GNU/Linux AND KDE from Debian themselves.
KDE is like the tail on the cat, helping it jump and keep its balance, whereas GNU is the legs, and Linux the heart. Chop a cat's legs off and it ain't going anywhere, no matter how fancy a tail it has. Chop it's tail off and its just clumsy.
Your point about the Qt license is well made, but Corel can avoid trouble by making KDE a seamless, albeit non-free (in the GPL sense) add-on.
Even Debian offers non-free code, they just keep it separate from GNU/Linux.
As for the O/S being tightly bound to the kernel, this is simply not true (and the biggest reaason to separate the names of the two): the core GNU O/S code runs on either a Linux or HURD kernel. Furthermore, you could replace the GNU O/S code with equivalent BSD code and have a BSD/Linux system. From a technical point of view, there is a good deal of sense to such nomenclature.
We've just become accustomed to thinking that the distribtion bundler has done more "work" and thus deserves greater "credit" than the source of much of the code common to all distributions.
Finally, its the Free Software Foundation and not the free software federation. They do not fight internally: ESR (and open source) are quite distinct from RMS (and free software). The GNU Project releases code when it's damn good and ready, and of extremely high quality, I might add. The HURD is very much a worthwhile, and difficult, project, though I suppose only a hacker would appreciate the inherent beauty of it.
RMS's views are strong, and I don't agree with all of them (in particular that any distribution carry a GNU/Linux moniker), but they serve a very useful function of setting the standard by which all compromise must be measured. And yes, this necessarily means that they might not always be practical or convenient.
Rip out the window manager and you have a fully functioning O/S.
Rip out the GNU code, and it all pretty much falls apart.
Of course, I would prefer that the GNU Project actually put together a distribution that could be called GNU/Linux (Debian's is close to this), with little justification for challenging that name. However, if they did, it would effectively be the common base for all other distros in that you could layer them over that base code. This happens now, in a virtual sense, hence the argument to call any distribution GNU/Linux.
I don't particularly think that argument is very strong, without a tangible GNU/Linux base, but I do think that if Corel is to build on an existing distribution that uses the GNU/Linux monkier, it should be retained.
While I generally think that the distribution provider gets to pick the distribution name, given that Corel's distribution will be based on Debian/GNU Linux, I would STRONGLY encourage Corel to name this distribution Corel GNU/Linux.
The justification for the GNU moniker comes from (a) RMS's reasoning that the operating system is clearly distinct from the kernel, (b) the plethora of GNU O/S software common to all distributions, and (c) being based on a distribution that has chosen to include GNU in it's name.
I've been generally reluctant to call ANY Linux distribution GNU/Linux (unless that is what the distributer calls it, or it was produced under the auspiscies of the GNU Project), though I do think that RMS's technical arguments for doing so are sound. Furthermore, the GNU Project's championship of free software should get a bit more exposure in all this Linux hysteria.
So, while I think that Corel does not have to call its distribution Corel GNU/Linux, I would very much like if they did.
I was under the presumption that the maximum penalty for (possession|traffiking, I don't know which) in pot was LIFE imprisonment. Someone in Ontario was facing this very bleak future for selling hemp seeds.
It is VERY difficult to compete with the coloquial moniker for something, even when the coloquial use is ambiguous and requires context to resolve the ambiguity. Witness that it is normal to call both the Linux kernel AND a complete O/S distribution Linux. Natural language is like that. I think that RMS is wasting his breath pushing the "GNU/Linux" moniker and his efforts would be better spent elsewhere.
That said, as Linux O/S distributions and other "Open Source" software become more popular, it WILL be important to distinguish between free (in the GPL sense) and not-so-free software. Here's where the GNU moniker, as an adjective, can come in handy.
I've proposed that the FSF authorize the use of the GNU moniker (trademark it, damn it! With an appropriate free license for use) to mean: that whose existence was facilitated by GPL software, and that is entirely consistent with the GPL.
Thus, "GNU Linux" (and drop the/) would be any linux distro that the FSF has blessed with the GNU moniker. Trust me, there WILL be less-free (in the FSF's opinion) Linux distros out there.
Of course, because the Linux kernel is GPL, it is the "GNU Linux" kernel, by definition.
Try getting a permit to TRANSPORT those firearms to the area where you intend to hunt.
You can't use them for self-defense, of course.
Yes, I am a Canadian (happily working in the free U.S. of A. with all it's quirks, pimples, and lower taxes). To each his own, I suppose.
If you think that a majority choice to strip a minority of its citizenship or other rights is a good thing, I suggest you be transformed to a Jew in the midst of Nazi Germany.
The first step will be to try and make programming illegal without a license, because of all the "harm" a badly/malicously written program can do on a networked system.
Yeah, though they claim it costs $105/day to run that site ?!? (in a blub to solicit more money) I wonder if that includes the cost of people to coordinate and handle things? If it's the cost of the site alone, someone's getting taken for a ride.
Until someone can read my mind without my intervention, my thoughts are mine, I'm afraid, as it would harm me (in terms of the time and effort it would take) to communicate them.
But the issue here really is about patent protection: should it be possible to make an idea public knowledge and simultaneously forbid others from using it without license?
Contractarians would say no: you can't restrict me without my consent, though, in practice. Your recourse is to not make the idea known to me. In practice of course, we accept all sorts of restrictions upon ourselves in exchange for others accepting similar restrictions upon THEMselves. We generally don't kill others because we don't want to be killed ourselves, to put it bluntly: killing is "wrong" because we think that "being killed" is wrong.
There are two counter arguments: First, that without protection against copying an idea, there is no incentive to innovate since the opportunity to profit is lost. Second, some ideas can't be exploited while keeping them secret (i.e. once shown, the idea is obvious).
But the key here is "copying". Patent protection not only provides recourse against unauthorized copying, it also prohibits indepedenendent reinvention. And, once the idea has been applied, reverse engineering may be trivial. This argument applies strongly in the pharmeceutical trade, where a new drug can cost millions to develop, but be copied easily.
Yes, the inventor is entitled to profit from his invention, and yes, he may require users to agree to indemnify him against their unauthorized copying. A life-saving invention that is arbitrarly expensive and reserved for the very rich is still better than it not being invented.
However, that's not what patents do. They restrict independent rediscovery, made all the more desirable when a solution to a problem is demonstrated to exist (by the first inventor), but kept secret and expensive.
Again, the argument is, "but we can't profit AND keep it secret". Tough noogies, I say.
So, restrictions against COPYING are fine, since ideas are properly the property of the thinker, however, restrictions against reinvention are not. Furthermore, the burden of proof should rest on the accuser, instead of the defendant, in such a dispute.
They may be proper, but I still consider them poor form.
"That is a rule with which I shall not put up" is the proper form of Churchill's basterdization, and not what he proposes. Though it is amazing how prevelant split infinitives and ending sentences with prepositions has become.
IOW, it works, and does what it does very very well.
Of course, I design and develop software for a living, and things like perl, sed, awk, grep, emacs, and others are the screwdrivers and pliers that get me through my day. While they've been ported to run in a DOS box under '95, they don't run as well. For some reason, they're real slow to boot (I think I know why).
Unfortunately, my target debugger and compiler run under DOS/Windows, so I'm stuck dual booting. But, even with the hassles this causes, it's STILL FASTER for me to reboot into Linux, run a 1000 file grep, and reboot back into '95.
Linux works. Windows doesn't. 'Nuff said.
Maybe at a monkey school like yours!
on
Salary Histories
·
· Score: 1
I remember an undergrad course I took, titled "Analysis of Algorithms". Innocent enough. Each week we had an assignment with about 3 to 4 problems, and one was ALWAYS an unsolved problem in computer science (i.e. is Ford-Johnson sort optimal for some 'n').
We were allowed to work together on assignments, but had to produce as many "sufficiently distinct and novel" approaches to each problem as there were collaborators.
The class very quickly dwindled to four students, one of whom solved almost all the unknown problems. He later went on to publish some interesting papers on sorting. His Masters' thesis defense was a laugh: he'd make a statement, one of his examiners would challenge it as a known unsolved problem, and he'd reply (in very halting 'VietnamEnglish', "No. Is obvious. I prove. Last month." and produce a reprint of the paper he just had published.
Needless to say, all four of us went on to get scholarships galore, graduate degrees, and bigger and better things.
Another seminar graduate course had us producing typed reports once a week, with 1/2 mark out of 10 taken off for spelling mistakes. I remember one assignment where I lost a 1/4 mark because an 'r' had a gap in the hook due to a faulty film ribbon used in the multiwriter printer. "You didn't proofread!," was the profs retort.
These courses weren't typical, of course, but some of us masochists loved them as they let us clearly differentiate ourselves from the mob.
Thanks to profs Bui and Suen at Concordia University, Montreal Canada, c. 1982-84, for all their encouragement and pushing us to our limits.
There appears to be some confusion about Katz's denegration of "Walls", after all aren't some Walls (privacy, etc.) good things?
The Walls Katz writes about are those imposed by others (governments, and corporations with government-backed enforcement) to divide and conquer us and not THOSE that we impose ourselves to keep others away from us.
"Duh, I put up a, er, 'wall' to keep people ouata my house, yeah, dat's it. Dat's good. Walls is good. So, when the gummint puts up walls dat must be good too, eh?"
I guess it's true: half the people you meet (ignoring the difference between mean, mode, and median) are dumber than you, assuming you're of average intelligence. IOW, not only are most people stupid, they're stupider than you think. Explains a long of things in a democracy, no?
Ya know, if you max out your credit card(s), you can live pretty high on the hog for a while... until the bills come.
This is exactly what socialist democracies like Canada and New Zealand (and others) do, with the exception, that the government maxes out each of the residents' "credit cards", and if your a lucky citizen, you might get something back for the charge, less "administrative government expenses", of course.
I was born, and lived in, Canada from 1961 to 1997, and I saw the rise of socialized this and that, to see it all crumble pretty badly by the time I left. I paid the approximate equivalent of US$23k income tax on an income of US$43k, supporting a non-employed spouse and child, as well as myself. Add to that 8% provincial and 7% federal "value added" taxes on about everything you buy. Take away all the supposed social benefits that got "clawed back" because I was a "high income" earner.
Socialist democracies are unsustainable in the long term, regardless of how seductive the idea appears in the short term.
Pointing to the failings and corruption within the U.S. government (of which there are many), blaming them on a capitalist economic system, and suggesting that some kind of socialist democracy is the answer is either very short-sighted thinking, or an attempt to stage a communist-style coup d'etat.
Time and time again, I've seen (from the inside, at various companies that I've worked for in the past) software that was released without adequate testing. Instead of stressing those parts that are particularly sensitive, and seeing if they break, all too often, some "magic" number of tests are run, under normal operating conditions, to see if it works. When it passes that number of tests, it's time to "ship it".
When a problem is encountered in the field, those same tests are run in field conditions (which, admitedly may not permit stress testing), to see "how many" fail.
Theories are advanced, one is picked by committee, and a change is made. If the same tests now succeed more often, progress is declared to have been made, and again, we "ship it".
Of course, the test sample is usually too small to come to any meaningful conclusions regarding improvement, but when the customer complains that the problem remains, we can say, "But look, we did 'stuff' and it got 'better'".
The advantage (?) of this approach exists purely in the mind of marketting droids for they can promise "a fix" by a certain arbitrary date.
On a related note, I've seen people in the business rewarded for "trying hard", instead of producing robust software, and staking their reputations on it.
Root cause analysis is often a tedious, time consuming, and painful task, often with little to show for the effort until the very end. (Somehow, "we don't know what it is yet, but we know what it isn't" doesn't count as progress. However, it is the ONLY way I know of to find and kill bugs once and for all. Unfortunately, this approach means that, "yes, we WILL find it, but it might take a day, or it might take a month." This isn't good enough for the marketting droids who's hollow delivery promises have come home to roost.
Recall? I guess, but it's tough to recall a solar powered weather station at the top of a mountain that requires access by helicopter. Multiply by several hundred.
Of course the silver lining to this particular cloud is that open source software results in such high quality that proprietary software companies will have to eventually meet the same high standard, one way or another.
Slashdot Mirror
Supposedly they were teased because they were homosexuals.
Certainly, this doesn't justify murder in response.
However, what everyone seams to be missing is that justification is NOT required: push someone enough and they might kill in response.
You don't do this for the same reason that you don't cross the street with your eyes closed.
This is the first bit of sense regarding ths matter I've seen.
Those that tease, taunt. and torture, should expect that some of the people they treat this way will react by killing others.
If you constantly beat a dog with a stick, the next time he sees a child with a similar strick, he will likely attack her.
Does this mean that the killers are blameless? No, but those that do things that are "wrong, but silently accepted" should be held accountable as well for encouraging others to become unstable.
This isn't the same as admonishing people to "walk on eggshells". Just because something one does may annoy somebody doesn't make it wrong. However, if one would not wish to be on the receiving end of what they dish out, they should be held accountable for the consequences.
Original post: the major players here are someone who didn't bother or didn't succeed to instill a sense of morality
My response: Er, no. The major players here are those whose teasing and taunts drove someone to lash back.
Your analysis: All right, you're saying that the real perpetrators of this crime are the various Jocks, Soshs, and Primadonnas of the Colombine HS social structure because they picked on some kids they thought were weird.
No. I used the term major players, to refer to those who could prevent such a tragedy. I disagree with the person who's post I followed up about just who the major players are. If I meant to write about the alleged (remember, even they are innocent until proven guilty) perpetrators of the crime, I would have used different words.
Since your understanding of what I wrote is in error, your further analysis is invalid.
the major players here are someone who didn't bother or didn't succeed to instill a sense of morality
Er, no. The major players here are those whose teasing and taunts drove someone to lash back.
What sense is there to instill a sense of morality and decency in someone if OTHERS will not act accordingly?
"Do unto others as you would have them do unto you" becomes perverted into "Do unto others as they have done unto you".
When some people disregard other's freedom to enjoy so-called alternate lifestyles, they should not be surprised when their freedoms are totally disregarded in return.
The overblown response of murder to teasing is perfectly in-line with the reasoning that the best response is a strong one (and why the U.S. floundered in Vietnam and won in the Gulf).
Is such a reaction right or justified? Certainly not, but the risk of it happening is to be expected. It is truely unfortunate that the response is directed so poorly at the instigators, but then again, who among is us innocent when we allow harrassment to persist?
The best premptive measures are those that serve to nip harrassment in the bud, and thus not giving rise to any form of retaliation, overblown or otherwise.
While I do not advocate taking the law into one's own hands to settle a dispute, neither do I care much when the rights of those who are disrespectful of others are themselves trampled upon. I do wish, however, that when such an overblown response occurs, it occur with greater regard for who the recipient is. Sadly, I do not ever expect this to be the case.
Its not that simple: layering KDE on top of a Debian distribution requires identifying HOW to effect such layering: where to put stuff, etc.
While KDE is not free, the information that makes the layering seamless can be.
Corel's distribution necessarily must be more than just Debian GNU/Linux + KDE. If not, then you could just get Debian GNU/Linux AND KDE from Debian themselves.
KDE is like the tail on the cat, helping it jump and keep its balance, whereas GNU is the legs, and Linux the heart. Chop a cat's legs off and it ain't going anywhere, no matter how fancy a tail it has. Chop it's tail off and its just clumsy.
Your point about the Qt license is well made, but Corel can avoid trouble by making KDE a seamless, albeit non-free (in the GPL sense) add-on.
Even Debian offers non-free code, they just keep it separate from GNU/Linux.
As for the O/S being tightly bound to the kernel, this is simply not true (and the biggest reaason to separate the names of the two): the core GNU O/S code runs on either a Linux or HURD kernel. Furthermore, you could replace the GNU O/S code with equivalent BSD code and have a BSD/Linux system. From a technical point of view, there is a good deal of sense to such nomenclature.
We've just become accustomed to thinking that the distribtion bundler has done more "work" and thus deserves greater "credit" than the source of much of the code common to all distributions.
Finally, its the Free Software Foundation and not the free software federation. They do not fight internally: ESR (and open source) are quite distinct from RMS (and free software). The GNU Project releases code when it's damn good and ready, and of extremely high quality, I might add. The HURD is very much a worthwhile, and difficult, project, though I suppose only a hacker would appreciate the inherent beauty of it.
RMS's views are strong, and I don't agree with all of them (in particular that any distribution carry a GNU/Linux moniker), but they serve a very useful function of setting the standard by which all compromise must be measured. And yes, this necessarily means that they might not always be practical or convenient.
Why?
Rip out the window manager and you have a fully functioning O/S.
Rip out the GNU code, and it all pretty much falls apart.
Of course, I would prefer that the GNU Project actually put together a distribution that could be called GNU/Linux (Debian's is close to this), with little justification for challenging that name. However, if they did, it would effectively be the common base for all other distros in that you could layer them over that base code. This happens now, in a virtual sense, hence the argument to call any distribution GNU/Linux.
I don't particularly think that argument is very strong, without a tangible GNU/Linux base, but I do think that if Corel is to build on an existing distribution that uses the GNU/Linux monkier, it should be retained.
While I generally think that the distribution provider gets to pick the distribution name, given that Corel's distribution will be based on Debian/GNU Linux, I would STRONGLY encourage Corel to name this distribution Corel GNU/Linux.
The justification for the GNU moniker comes from (a) RMS's reasoning that the operating system is clearly distinct from the kernel, (b) the plethora of GNU O/S software common to all distributions, and (c) being based on a distribution that has chosen to include GNU in it's name.
I've been generally reluctant to call ANY Linux distribution GNU/Linux (unless that is what the distributer calls it, or it was produced under the auspiscies of the GNU Project), though I do think that RMS's technical arguments for doing so are sound. Furthermore, the GNU Project's championship of free software should get a bit more exposure in all this Linux hysteria.
So, while I think that Corel does not have to call its distribution Corel GNU/Linux, I would very much like if they did.
SCO just doesn't get it.
It appears that Red Hat is doing quite well, thank you, with their business model.
When did this happen?
I was under the presumption that the maximum penalty for (possession|traffiking, I don't know which) in pot was LIFE imprisonment. Someone in Ontario was facing this very bleak future for selling hemp seeds.
Whatever happened to the notion of a "jury of one's PEERS"?
I've suggested something similar to the FSF.
/) would be any linux distro that the FSF has blessed with the GNU moniker. Trust me, there WILL be less-free (in the FSF's opinion) Linux distros out there.
It is VERY difficult to compete with the coloquial moniker for something, even when the coloquial use is ambiguous and requires context to resolve the ambiguity. Witness that it is normal to call both the Linux kernel AND a complete O/S distribution Linux. Natural language is like that. I think that RMS is wasting his breath pushing the "GNU/Linux" moniker and his efforts would be better spent elsewhere.
That said, as Linux O/S distributions and other "Open Source" software become more popular, it WILL be important to distinguish between free (in the GPL sense) and not-so-free software. Here's where the GNU moniker, as an adjective, can come in handy.
I've proposed that the FSF authorize the use of the GNU moniker (trademark it, damn it! With an appropriate free license for use) to mean: that whose existence was facilitated by GPL software, and that is entirely consistent with the GPL.
Thus, "GNU Linux" (and drop the
Of course, because the Linux kernel is GPL, it is the "GNU Linux" kernel, by definition.
Try getting a permit to TRANSPORT those firearms to the area where you intend to hunt.
You can't use them for self-defense, of course.
Yes, I am a Canadian (happily working in the free U.S. of A. with all it's quirks, pimples, and lower taxes). To each his own, I suppose.
If you think that a majority choice to strip a minority of its citizenship or other rights is a good thing, I suggest you be transformed to a Jew in the midst of Nazi Germany.
The first step will be to try and make programming illegal without a license, because of all the "harm" a badly/malicously written program can do on a networked system.
Yeah, though they claim it costs $105/day to run that site ?!? (in a blub to solicit more money) I wonder if that includes the cost of people to coordinate and handle things? If it's the cost of the site alone, someone's getting taken for a ride.
The Canadaian government has just about outlawed non-police possession of firearms.
You don't "need" to hunt - you can buy food, ergo no need for a shotgun. You don't "need" to defend yourself - that's what the cops are for.
The day my taxes pay for a rent-a-cop by my side up to 24 hours a day, I may consider moving back there.
Until someone can read my mind without my intervention, my thoughts are mine, I'm afraid, as it would harm me (in terms of the time and effort it would take) to communicate them.
But the issue here really is about patent protection: should it be possible to make an idea public knowledge and simultaneously forbid others from using it without license?
Contractarians would say no: you can't restrict me without my consent, though, in practice. Your recourse is to not make the idea known to me. In practice of course, we accept all sorts of restrictions upon ourselves in exchange for others accepting similar restrictions upon THEMselves. We generally don't kill others because we don't want to be killed ourselves, to put it bluntly: killing is "wrong" because we think that "being killed" is wrong.
There are two counter arguments: First, that without protection against copying an idea, there is no incentive to innovate since the opportunity to profit is lost. Second, some ideas can't be exploited while keeping them secret (i.e. once shown, the idea is obvious).
But the key here is "copying". Patent protection not only provides recourse against unauthorized copying, it also prohibits indepedenendent reinvention. And, once the idea has been applied, reverse engineering may be trivial. This argument applies strongly in the pharmeceutical trade, where a new drug can cost millions to develop, but be copied easily.
Yes, the inventor is entitled to profit from his invention, and yes, he may require users to agree to indemnify him against their unauthorized copying. A life-saving invention that is arbitrarly expensive and reserved for the very rich is still better than it not being invented.
However, that's not what patents do. They restrict independent rediscovery, made all the more desirable when a solution to a problem is demonstrated to exist (by the first inventor), but kept secret and expensive.
Again, the argument is, "but we can't profit AND keep it secret". Tough noogies, I say.
So, restrictions against COPYING are fine, since ideas are properly the property of the thinker, however, restrictions against reinvention are not. Furthermore, the burden of proof should rest on the accuser, instead of the defendant, in such a dispute.
I'm sorry, I hate split infinitives.
They may be proper, but I still consider them poor form.
"That is a rule with which I shall not put up" is the proper form of Churchill's basterdization, and not what he proposes. Though it is amazing how prevelant split infinitives and ending sentences with prepositions has become.
In my experience, NT's TCP/IP stack is about the WORST implementation I've seen. Well, second worse, 95's takes the cake.
The only time I've seen Linux crash is when I've been monkeying with the kernel, or various drivers, or had a hardware problem.
With friends like that, who needs enemies?
I use Linux because it DOESN'T CRASH ON ME!
IOW, it works, and does what it does very very well.
Of course, I design and develop software for a living, and things like perl, sed, awk, grep, emacs, and others are the screwdrivers and pliers that get me through my day. While they've been ported to run in a DOS box under '95, they don't run as well. For some reason, they're real slow to boot (I think I know why).
Unfortunately, my target debugger and compiler run under DOS/Windows, so I'm stuck dual booting. But, even with the hassles this causes, it's STILL FASTER for me to reboot into Linux, run a 1000 file grep, and reboot back into '95.
Linux works. Windows doesn't. 'Nuff said.
I remember an undergrad course I took, titled "Analysis of Algorithms". Innocent enough. Each week we had an assignment with about 3 to 4 problems, and one was ALWAYS an unsolved problem in computer science (i.e. is Ford-Johnson sort optimal for some 'n').
We were allowed to work together on assignments, but had to produce as many "sufficiently distinct and novel" approaches to each problem as there were collaborators.
The class very quickly dwindled to four students, one of whom solved almost all the unknown problems. He later went on to publish some interesting papers on sorting. His Masters' thesis defense was a laugh: he'd make a statement, one of his examiners would challenge it as a known unsolved problem, and he'd reply (in very halting 'VietnamEnglish', "No. Is obvious. I prove. Last month." and produce a reprint of the paper he just had published.
Needless to say, all four of us went on to get scholarships galore, graduate degrees, and bigger and better things.
Another seminar graduate course had us producing typed reports once a week, with 1/2 mark out of 10 taken off for spelling mistakes. I remember one assignment where I lost a 1/4 mark because an 'r' had a gap in the hook due to a faulty film ribbon used in the multiwriter printer. "You didn't proofread!," was the profs retort.
These courses weren't typical, of course, but some of us masochists loved them as they let us clearly differentiate ourselves from the mob.
Thanks to profs Bui and Suen at Concordia University, Montreal Canada, c. 1982-84, for all their encouragement and pushing us to our limits.
Dear IRS:
I can not tell you how much I make because my employer threatened to fire me if I do.
Sincerely,
Gagged.
Dear Gagged:
If you do not tell us how much you make, we will jail you.
Sincerely,
IRS
Take your pick.
There appears to be some confusion about Katz's denegration of "Walls", after all aren't some Walls (privacy, etc.) good things?
The Walls Katz writes about are those imposed by others (governments, and corporations with government-backed enforcement) to divide and conquer us and not THOSE that we impose ourselves to keep others away from us.
"Duh, I put up a, er, 'wall' to keep people ouata my house, yeah, dat's it. Dat's good. Walls is good. So, when the gummint puts up walls dat must be good too, eh?"
I guess it's true: half the people you meet (ignoring the difference between mean, mode, and median) are dumber than you, assuming you're of average intelligence. IOW, not only are most people stupid, they're stupider than you think. Explains a long of things in a democracy, no?
Ya know, if you max out your credit card(s), you can live pretty high on the hog for a while... until the bills come.
This is exactly what socialist democracies like Canada and New Zealand (and others) do, with the exception, that the government maxes out each of the residents' "credit cards", and if your a lucky citizen, you might get something back for the charge, less "administrative government expenses", of course.
I was born, and lived in, Canada from 1961 to 1997, and I saw the rise of socialized this and that, to see it all crumble pretty badly by the time I left. I paid the approximate equivalent of US$23k income tax on an income of US$43k, supporting a non-employed spouse and child, as well as myself. Add to that 8% provincial and 7% federal "value added" taxes on about everything you buy. Take away all the supposed social benefits that got "clawed back" because I was a "high income" earner.
Socialist democracies are unsustainable in the long term, regardless of how seductive the idea appears in the short term.
Pointing to the failings and corruption within the U.S. government (of which there are many), blaming them on a capitalist economic system, and suggesting that some kind of socialist democracy is the answer is either very short-sighted thinking, or an attempt to stage a communist-style coup d'etat.
Time and time again, I've seen (from the inside, at various companies that I've worked for in the past) software that was released without adequate testing. Instead of stressing those parts that are particularly sensitive, and seeing if they break, all too often, some "magic" number of tests are run, under normal operating conditions, to see if it works. When it passes that number of tests, it's time to "ship it".
When a problem is encountered in the field, those same tests are run in field conditions (which, admitedly may not permit stress testing), to see "how many" fail.
Theories are advanced, one is picked by committee, and a change is made. If the same tests now succeed more often, progress is declared to have been made, and again, we "ship it".
Of course, the test sample is usually too small to come to any meaningful conclusions regarding improvement, but when the customer complains that the problem remains, we can say, "But look, we did 'stuff' and it got 'better'".
The advantage (?) of this approach exists purely in the mind of marketting droids for they can promise "a fix" by a certain arbitrary date.
On a related note, I've seen people in the business rewarded for "trying hard", instead of producing robust software, and staking their reputations on it.
Root cause analysis is often a tedious, time consuming, and painful task, often with little to show for the effort until the very end. (Somehow, "we don't know what it is yet, but we know what it isn't" doesn't count as progress. However, it is the ONLY way I know of to find and kill bugs once and for all. Unfortunately, this approach means that, "yes, we WILL find it, but it might take a day, or it might take a month." This isn't good enough for the marketting droids who's hollow delivery promises have come home to roost.
Recall? I guess, but it's tough to recall a solar powered weather station at the top of a mountain that requires access by helicopter. Multiply by several hundred.
Of course the silver lining to this particular cloud is that open source software results in such high quality that proprietary software companies will have to eventually meet the same high standard, one way or another.