The quote was not sufficiently distinct to be recognized as a Simpson's quote. Lots of people say that, and with that exact phrasing. It is not reasonable to assume that people would recognize it as a Simpson's quote, even if they HAD seen that episode. It's like saying "the" and calling it a Simpsons quote becasue there was that one episode where Marge said "the" in a sentence.
If it was me, I would have been charged with perjury and put in jail...
If it was you, you wouldn't have been on trial for having consentual sexual relations with people, since that's not a crime (it can be grounds for divorce, but that's not a crime). Did Clinton lie under oath? Yes, absolutely. Was the trial he was being put under at the time a legitimate one? No, absolutely not. Both parties were full of idiocy in the rhetoric surrounding that incident.
SOME of us here work with computers, so we know that sometimes the computers make mistakes. Nah. We know that the people who programmed them make mistakes, and the computer is only as trustable as its programmer. The computer itself just does whatever it's told, even if its a bug that alters the totals in a vote counting machine.
" Ever since I can remember, I've been a neo-something: a neo-Marxist, a neo-Trotskyist, a neo-liberal, a neo-conservative; in religion a neo-orthodox even while I was a neo-Trotskyist and a neo-Marxist. I'm going to end up a neo- that's all, neo dash nothing."
I mean, we had uncounted ballot boxes floating in San Fransisco bay, but all the media could talk about was how Jeb Bush deliberately forced people to use Nazi-inspired punched ballots just so his brother could win.
Your vote wasn't close enough for it to matter. If it was, then those floating boxes would have been VERY relevant and had just as much media attention. ANY votes lost during handling constitute a disenfranchisement of voters.
But really, it doesn't require much more than an IQ of 70 to learn how to use a punch-card ballot -- AND make sure the chads are completely removed...
It takes luck to ensure that nothing happens to the card after you punched it. Once it leaves your hand, it is not immediately fed into a card reader in front of your eyes. The chad system is fragile enough that hanging chads, extra punches, and stray chads from neighboring cards can be introduced during shipping and handling.
Do not assume that all fouled punch ballots got that way because of the voters. A fragile record-keeping system like that is completely unacceptable when it needs to be transported before counting.
And, punchcard readers can introduce additional folding, mutilating, or spindling when they process the card. So on a recount vote, the record of who voted for what has been altered by the damage caused to sone of the cards on the first run through the machines. Recounts are another reason it is unacceptable to have a voting system with a fragile record-keeping ability. The ballot has to have the endurace to survive the counting process without introducing any changes.
Electronic voting machines are shown to dramatically reduce the error rate in elections.
The error rate on the part of voters or counters? If it's only on the part of voters, then why not just use an electronic interface who's sole purpose is to print out a filled-out paper ballot that gets dropped in a box and manually counted later?
The news media played this one down, making the excuse that it is not yet election season, so people can't be expected to know!
But that is a perfectly legitimate excuse! It doesn't take an entire year to learn about the candidates and decide who your choice will be. This early on, people claiming they plan to run might still change their minds before the relevant vote occurs. So There's no reason to bother getting started this early. It's like starting your Christmas shopping in March.
To be precise, we have a person suing a restaurant because it sells a product that they intend for you to put in your mouth despite the fact that it is hot enough to cause 3rd degree burns. They do this despite the fact that they KNOW FROM EXPERIENCE that their actions will cause people to be injured. Excusing the restaurant because "people should know coffee is hot" is tantamount to excusing the presence of dangerous amounts of arsenic in their fries, because "people should know that fried food is unhealthy." Note that in the case you refer to, the restaurant was NOT sued until they had repeatedly refused to assist the burned woman in paying her medical expenses.
To be precise, she didn't really spill the coffee on herself either - the spill occurred during the transfer of the cup from the employee's hand to hers. It was not clear who was more at fault, her or the employee. So the line of argument that it was all her fault doesn't hold water even without asking if she should have expected it to be that hot. It's like the situation where a car accident happens and the insurance companies have to asses percentage of blame to the two parties. It's often NOT 100% one person's fault. McDonalds' lawyers chose to characterize the incident as "spilling coffee on herself" in the hope that this would provoke exactly the reaction it did from the peanut gallery.
I also have to point out that if this electronic voting were being taken care of by a private company,
You mean like Diebold?
these kinds of problems would have a more direct target for blame.
Already proven false. Diebold does have these problems, and the people who have been pointing them out have been ignored because the public aren't Diebold's customers, the politicians are.
I agree. If you cannot figure out how to punch the chad out, then your vote should not count.
And if an additional chad falls out during transport and handling because the cards are fragile, your vote shouldn't count either, since you were obviously dumb enough to vote for two people, right? And if a neighboring chad from the adjacent card in the stack gets pressed into the hole on your card and stuck there that must mean you are too stupid to punch your chad, right? Face it, the punchcard system has huge technical problems that have nothing to do with voter stupidity. It is ignorance to assume that after being stacked together, transported across the city, placed into the reader, and fed through the machine, that all fouled cards the reader rejects must have gotten that way because the voter submitted the card that way. The other bad ballot, the butterfly system, was bad because the holes were a seperate piece from the text labels that described them, and when the two were bound together they didn't always line up very well, sometimes making it ambiguous which hole was supposed to go with which name (And, NO you can't just count holes and assume for example that the third hole must be for the third name, because there were more holes than names - some holes were 'dummy' holes (printing a name only for every odd-numbered hole). Once the binding is misaligned you lost your only indicator as to which hole was which.)
It should be that simple. If you are not smart enough to figure out how to vote, why would you be smart enough to choose our leaders?
I'm all for poll tests (for everyone), some basic questions, like what is the 1st amendment to the Constitution or how many amendments are in the Bill of Rights. You wouldn't let children vote on how to spend their parents money would you?
I'd be more concerned with whether people paid attention to what the candidate's platforms are. I despise blind-faith party-line voting. Have a test with candidates shown on the left, and small paragraphs shown on the right where each candidate describes one or two key points of their platform The test is to match up the right candidates with their paragraph statement. If you can't, then you are an uninformed voter and are ruining democracy by your participation. even if you do know the first amendment.
But there's a much more direct correlation between how bug-free a program is and how much effort is invested in quality control, during both the design and implementation phases.
No - the correlation is between bug-free and the relative DIFFERENCE between how much quality control there was and how much quality control there needed to be. As you say, if you increase the complexity you increase the amount of QC needed. Thus two programs with exactly the same amount of QC will have different levels of bugginess if one was more complex than the other.
Depends upon the computer system. Embedded systems (such as those in your microwave, or those that control massive industrial processes) only do one thing for their entire life-cycle, and they can be tested for most possible input conditions.
I think this supports my point well, given that those types of systems aren't as buggy as desktop software. There appears to be a correlation between how multipurpose a piece of software is and how buggy it is.
Actually, I have read about a Linux anti-virus program, but I was so concerned that I forgot the name.:)
I know of that one too (and can't remember the name either), but it exists NOT to protect Linux from viruses. It's a virus scanner to help prevent a linux box from propigating viruses for other OS's if they travel *through* the linux box. It was meant for cases where (for example) a linux box is serving mail to Windows machines.)
Check out the weather stats for Salt Lake City, UT [wunderground.com] (51" of snow a year).
I explicitly mentioned that places near mountains (for which Salt Lake City would qualify) are an exception. Ot at least I tried to. Upon re-reading my post I see that it was kind of confusing.
Many other large cities in the above mentioned states many not get that kind of snowfall, but still regularly have below freezing temperatures throughout winter (especially at night).
I don't consider it all that wintery if the temperature only goes below freezing at night. Stuff can still actually grow in weather like that.
Any user who does not patch daily and harms another due to not being patched should be punished. Here is how I think it should work....
Seeing as how there has been occurances in the past where a patch solved one problem but caused another, I don't think that would be fair. If installing a patch will break something you *need*, but not a lot of other people do, is it fair for them to force you to install it?
One problem with this is that, if you eliminate all the free and open software, you will effectively shut down most of the Internet, and that ain't gonna happen.
You assume the regulators will be competent enough to realize this. It's entirely possible that they hypocritically keep using the free software that runs the internet, either by artful dodging of the truth, or more likely, through sheer incompetence where they don't realize that's what they're doing, and attempts to tell them about it will be ignored as "irrelevant whining".
And yes, buildings don't fall (very often) and bridges don't collapse (most of the time), BUT the difference there is that the technologies and math used to design those constructs have been well understood for centuries.
I don't think that's the reason. I think the reason is that the tasks computers perform have a lot more variety than anything else. If you design a bridge, it has one purpose only - to carry a road - which has weight travelling across it. You aren't going to suddenly find next week that someone also wants to keep their recepies on the thing, and use it to pretend they're shooting space aliens with ray guns, and then use it to balance their checkbook.
Software breaks because the conditions that have to be tested for are almost infinite in variety. So the best the software engineers can do is try to test for certain classes of things they think could be relevant, and hope that all the uses peopel come up with for the computer fall under those classes of things that were tested against.
One of the things that worries me is things like this:
I work on the staff of a university. At the student union, they have a free access wireless LAN that covers the building and lets anyone with a laptop hook up to the campus network, get a DHCP-assigned IP and run with it, no questions asked, no login required. When the SoBIG virus was running amok a week or two back, they started putting up signs that said anyone using that wireless network MUST have antivirus software that blocks e-mail viruses installed and up-to date or they don't have the rights to use the network. (followed by a URL for where they could get such software from the campus IT department) Now, aside from how on earth they thought they could enforce this, I was a bit worried about what they would say about someone not using Microsoft. Would they be dumb enough (assuming they could even find out) to say that someone using a Linux laptop was violating the rule and being "less" safe than someone using Windows with the anti-virus software installed? Would they understand that that software doesn't exist on Linux only because there's no need for it (yet?).
Will the idiots in charge believe that a computer with antivirus software on it is always more 'trusted' than one without, even if the reason it lacks the software is because it is at zero risk since it's not running the OS that has the viruses?
But there's something to be said for a language that makes it simple to do the right thing, and requires a bit more explicitness and verbosity to do the wrong thing.
The problem is that people disagree strongly over which way is the "wrong thing". Some would say that the wasted CPU time constantly checking every array reference every time (such as happens in languages that enforce array bounds for you) is the "wrong thing". (Ideally, I think the programmer should be able to invoke array bounds checking when he wants to, with a compiler directive, and turn it off when he wants to. Then he could wrap the parts that process user's untrusted input with this directive, but leave it off the rest of the time for speed.)
Actually, I wouldn't blame C so much as the standard library. If there were a decent set of variable-sized array classes, safe string classes and so on included as part of the standard C library, we probably wouldn't see so many problems.
I agree there. Every time I had to program something that had to protect itself against buffer overflows, I got very annoyed at the lack of strn... functions in the library. For working with normal strings, there's lots of useful functions to play with: strstr, strtok, strchr, strdup, and so on, but only a select few of them have "n" versions, like strncpy, and strncmp. Why this is the case I don't know. But I'm sure it's part of the reason people don't check their bounds as much as they should. If you want to use the useful C-string library functions, you have to do so unsafely.
The safety of which you speak has nothing to do with how high-level the language is. In C, the decision not to check array bounds is not a matter of it being low level. It's a matter of preferring speed over coddling. Every array reference takes three times longer when you have bounds checking. The design model is that you only use the checking when you have to (like when filling a buffer with untrusted input). The fact that programmers don't do that is their fault, not the language's fault.
(That being said, I think it would be cool for a language to have a pragma that could turn bounds checking on and off when the programmer wants to, so a critical section of code that processes user input could operate with it on, but the rest of the code operate with it off.)
His point wasn't that they aren't capable enough to do it. His point was that the printer in need of fixing isn't *IN* India. You can't fix a printer on the other side of the world.
Well, yes... if you want to freeze to death during the rather unpleasant period known as "winter".
Arizona, New Mexico, Colorado, Utah, Oklahoma, Texas, Lousiana, Arkansas, Mississippi, Tennassee, Florida, Georgia, South Carolina. North Carolina. These are all places that only get snow on occasion (or near mountains). You have to live in the northern 1/3 of the country to really have wintery weather.
Slashdot Mirror
The quote was not sufficiently distinct to be recognized as a Simpson's quote. Lots of people say that, and with that exact phrasing. It is not reasonable to assume that people would recognize it as a Simpson's quote, even if they HAD seen that episode. It's like saying "the" and calling it a Simpsons quote becasue there was that one episode where Marge said "the" in a sentence.
If it was me, I would have been charged with perjury and put in jail...
If it was you, you wouldn't have been on trial for having consentual sexual relations with people, since that's not a crime (it can be grounds for divorce, but that's not a crime). Did Clinton lie under oath? Yes, absolutely. Was the trial he was being put under at the time a legitimate one? No, absolutely not. Both parties were full of idiocy in the rhetoric surrounding that incident.
SOME of us here work with computers, so we know that sometimes the computers make mistakes.
Nah. We know that the people who programmed them make mistakes, and the computer is only as trustable as its programmer. The computer itself just does whatever it's told, even if its a bug that alters the totals in a vote counting machine.
" Ever since I can remember, I've been a neo-something: a neo-Marxist, a neo-Trotskyist, a neo-liberal, a neo-conservative; in religion a neo-orthodox even while I was a neo-Trotskyist and a neo-Marxist. I'm going to end up a neo- that's all, neo dash nothing."
In that case, Neo-, I reccomend the blue pill.
I mean, we had uncounted ballot boxes floating in San Fransisco bay, but all the media could talk about was how Jeb Bush deliberately forced people to use Nazi-inspired punched ballots just so his brother could win.
Your vote wasn't close enough for it to matter. If it was, then those floating boxes would have been VERY relevant and had just as much media attention. ANY votes lost during handling constitute a disenfranchisement of voters.
But really, it doesn't require much more than an IQ of 70 to learn how to use a punch-card ballot -- AND make sure the chads are completely removed...
It takes luck to ensure that nothing happens to the card after you punched it. Once it leaves your hand, it is not immediately fed into a card reader in front of your eyes. The chad system is fragile enough that hanging chads, extra punches, and stray chads from neighboring cards can be introduced during shipping and handling.
Do not assume that all fouled punch ballots got that way because of the voters. A fragile record-keeping system like that is completely unacceptable when it needs to be transported before counting.
And, punchcard readers can introduce additional folding, mutilating, or spindling when they process the card. So on a recount vote, the record of who voted for what has been altered by the damage caused to sone of the cards on the first run through the machines. Recounts are another reason it is unacceptable to have a voting system with a fragile record-keeping ability. The ballot has to have the endurace to survive the counting process without introducing any changes.
Electronic voting machines are shown to dramatically reduce the error rate in elections.
The error rate on the part of voters or counters? If it's only on the part of voters, then why not just use an electronic interface who's sole purpose is to print out a filled-out paper ballot that gets dropped in a box and manually counted later?
The news media played this one down, making the excuse that it is not yet election season, so people can't be expected to know!
But that is a perfectly legitimate excuse! It doesn't take an entire year to learn about the candidates and decide who your choice will be. This early on, people claiming they plan to run might still change their minds before the relevant vote occurs. So There's no reason to bother getting started this early. It's like starting your Christmas shopping in March.
To be precise, we have a person suing a restaurant because it sells a product that they intend for you to put in your mouth despite the fact that it is hot enough to cause 3rd degree burns. They do this despite the fact that they KNOW FROM EXPERIENCE that their actions will cause people to be injured. Excusing the restaurant because "people should know coffee is hot" is tantamount to excusing the presence of dangerous amounts of arsenic in their fries, because "people should know that fried food is unhealthy." Note that in the case you refer to, the restaurant was NOT sued until they had repeatedly refused to assist the burned woman in paying her medical expenses.
To be precise, she didn't really spill the coffee on herself either - the spill occurred during the transfer of the cup from the employee's hand to hers. It was not clear who was more at fault, her or the employee. So the line of argument that it was all her fault doesn't hold water even without asking if she should have expected it to be that hot. It's like the situation where a car accident happens and the insurance companies have to asses percentage of blame to the two parties. It's often NOT 100% one person's fault. McDonalds' lawyers chose to characterize the incident as "spilling coffee on herself" in the hope that this would provoke exactly the reaction it did from the peanut gallery.
I also have to point out that if this electronic voting were being taken care of by a private company,
You mean like Diebold?
these kinds of problems would have a more direct target for blame.
Already proven false. Diebold does have these problems, and the people who have been pointing them out have been ignored because the public aren't Diebold's customers, the politicians are.
I agree. If you cannot figure out how to punch the chad out, then your vote should not count.
And if an additional chad falls out during transport and handling because the cards are fragile, your vote shouldn't count either, since you were obviously dumb enough to vote for two people, right? And if a neighboring chad from the adjacent card in the stack gets pressed into the hole on your card and stuck there that must mean you are too stupid to punch your chad, right? Face it, the punchcard system has huge technical problems that have nothing to do with voter stupidity. It is ignorance to assume that after being stacked together, transported across the city, placed into the reader, and fed through the machine, that all fouled cards the reader rejects must have gotten that way because the voter submitted the card that way. The other bad ballot, the butterfly system, was bad because the holes were a seperate piece from the text labels that described them, and when the two were bound together they didn't always line up very well, sometimes making it ambiguous which hole was supposed to go with which name (And, NO you can't just count holes and assume for example that the third hole must be for the third name, because there were more holes than names - some holes were 'dummy' holes (printing a name only for every odd-numbered hole). Once the binding is misaligned you lost your only indicator as to which hole was which.)
It should be that simple. If you are not smart enough to figure out how to vote, why would you be smart enough to choose our leaders?
I'm all for poll tests (for everyone), some basic questions, like what is the 1st amendment to the Constitution or how many amendments are in the Bill of Rights. You wouldn't let children vote on how to spend their parents money would you?
I'd be more concerned with whether people paid attention to what the candidate's platforms are. I despise blind-faith party-line voting. Have a test with candidates shown on the left, and small paragraphs shown on the right where each candidate describes one or two key points of their platform The test is to match up the right candidates with their paragraph statement. If you can't, then you are an uninformed voter and are ruining democracy by your participation. even if you do know the first amendment.
The general public will accept a solution that is good enough.
The problem is that the general public will also accept a solution that is NOT good enough.
But there's a much more direct correlation between how bug-free a program is and how much effort is invested in quality control, during both the design and implementation phases.
No - the correlation is between bug-free and the relative DIFFERENCE between how much quality control there was and how much quality control there needed to be. As you say, if you increase the complexity you increase the amount of QC needed. Thus two programs with exactly the same amount of QC will have different levels of bugginess if one was more complex than the other.
Depends upon the computer system. Embedded systems (such as those in your microwave, or those that control massive industrial processes) only do one thing for their entire life-cycle, and they can be tested for most possible input conditions.
I think this supports my point well, given that those types of systems aren't as buggy as desktop software. There appears to be a correlation between how multipurpose a piece of software is and how buggy it is.
Actually, I have read about a Linux anti-virus program, but I was so concerned that I forgot the name.
I know of that one too (and can't remember the name either), but it exists NOT to protect Linux from viruses. It's a virus scanner to help prevent a linux box from propigating viruses for other OS's if they travel *through* the linux box. It was meant for cases where (for example) a linux box is serving mail to Windows machines.)
Check out the weather stats for Salt Lake City, UT [wunderground.com] (51" of snow a year).
I explicitly mentioned that places near mountains (for which Salt Lake City would qualify) are an exception. Ot at least I tried to. Upon re-reading my post I see that it was kind of confusing.
Many other large cities in the above mentioned states many not get that kind of snowfall, but still regularly have below freezing temperatures throughout winter (especially at night).
I don't consider it all that wintery if the temperature only goes below freezing at night.
Stuff can still actually grow in weather like that.
Any user who does not patch daily and harms another due to not being patched should be punished. Here is how I think it should work....
Seeing as how there has been occurances in the past where a patch solved one problem but caused another, I don't think that would be fair. If installing a patch will break something you *need*, but not a lot of other people do, is it fair for them to force you to install it?
One problem with this is that, if you eliminate all the free and open software, you will effectively shut down most of the Internet, and that ain't gonna happen.
You assume the regulators will be competent enough to realize this. It's entirely possible that they hypocritically keep using the free software that runs the internet, either by artful dodging of the truth, or more likely, through sheer incompetence where they don't realize that's what they're doing, and attempts to tell them about it will be ignored as "irrelevant whining".
And yes, buildings don't fall (very often) and bridges don't collapse (most of the time), BUT the difference there is that the technologies and math used to design those constructs have been well understood for centuries.
I don't think that's the reason. I think the reason is that the tasks computers perform have a lot more variety than anything else. If you design a bridge, it has one purpose only - to carry a road - which has weight travelling across it. You aren't going to suddenly find next week that someone also wants to keep their recepies on the thing, and use it to pretend they're shooting space aliens with ray guns, and then use it to balance their checkbook.
Software breaks because the conditions that have to be tested for are almost infinite in variety. So the best the software engineers can do is try to test for certain classes of things they think could be relevant, and hope that all the uses peopel come up with for the computer fall under those classes of things that were tested against.
One of the things that worries me is things like this:
I work on the staff of a university. At the student union, they have a free access wireless LAN that covers the building and lets anyone with a laptop hook up to the campus network, get a DHCP-assigned IP and run with it, no questions asked, no login required. When the SoBIG virus was running amok a week or two back, they started putting up signs that said anyone using that wireless network MUST have antivirus software that blocks e-mail viruses installed and up-to date or they don't have the rights to use the network. (followed by a URL for where they could get such software from the campus IT department) Now, aside from how on earth they thought they could enforce this, I was a bit worried about what they would say about someone not using Microsoft. Would they be dumb enough (assuming they could even find out) to say that someone using a Linux laptop was violating the rule and being "less" safe than someone using Windows with the anti-virus software installed? Would they understand that that software doesn't exist on Linux only because there's no need for it (yet?).
Will the idiots in charge believe that a computer with antivirus software on it is always more 'trusted' than one without, even if the reason it lacks the software is because it is at zero risk since it's not running the OS that has the viruses?
They just might be that dumb.
But there's something to be said for a language that makes it simple to do the right thing, and requires a bit more explicitness and verbosity to do the wrong thing.
The problem is that people disagree strongly over which way is the "wrong thing". Some would say that the wasted CPU time constantly checking every array reference every time (such as happens in languages that enforce array bounds for you) is the "wrong thing". (Ideally, I think the programmer should be able to invoke array bounds checking when he wants to, with a compiler directive, and turn it off when he wants to. Then he could wrap the parts that process user's untrusted input with this directive, but leave it off the rest of the time for speed.)
Actually, I wouldn't blame C so much as the standard library. If there were a decent set of variable-sized array classes, safe string classes and so on included as part of the standard C library, we probably wouldn't see so many problems.
I agree there. Every time I had to program something that had to protect itself against buffer overflows, I got very annoyed at the lack of strn... functions in the library. For working with normal strings, there's lots of useful functions to play with: strstr, strtok, strchr, strdup, and so on, but only a select few of them have "n" versions, like strncpy, and strncmp. Why this is the case I don't know. But I'm sure it's part of the reason people don't check their bounds as much as they should. If you want to use the useful C-string library functions, you have to do so unsafely.
The safety of which you speak has nothing to do with how high-level the language is. In C, the decision not to check array bounds is not a matter of it being low level. It's a matter of preferring speed over coddling. Every array reference takes three times longer when you have bounds checking. The design model is that you only use the checking when you have to (like when filling a buffer with untrusted input). The fact that programmers don't do that is their fault, not the language's fault.
(That being said, I think it would be cool for a language to have a pragma that could turn bounds checking on and off when the programmer wants to,
so a critical section of code that processes user input could operate with it on, but the rest of the code operate with it off.)
His point wasn't that they aren't capable enough to do it. His point was that the printer in need of fixing isn't *IN* India. You can't fix a printer on the other side of the world.
Well, yes
Arizona, New Mexico, Colorado, Utah, Oklahoma, Texas, Lousiana, Arkansas, Mississippi, Tennassee, Florida, Georgia, South Carolina. North Carolina. These are all places that only get snow on occasion (or near mountains). You have to live in the northern 1/3 of the country to really have wintery weather.
Man, that sounds a lot like a case of familiarity breeding usability.
I never said it became more "usable". I said people got used to it. That's not the same thing.