I doubt that matters. It's not really law anyway, because it's open to wildly varying interpretations. Law is objective, anti-trust "law" is subjective.
Really, who would sue? Novell couldn't last that long in court without cashflow. However long they could last would probably just result in a settlement much less generous than that with Sun & MS.
So let's say a state or the feds sue. MS would just drag it out, and since there is no smoking gun, nothing will really happen. What if the states win? What do they do? Anything the states do against MS would just be hitting the stockholders. And by stockholders, I mean the retired people living on diversified investment income. The execs have already made the money from the "anticompetitive" behavior, probably much of it in terms of huge bonuses.
Anti trust law is a mess. It doesn't really accomplish anything. After all, they went after IBM, and what happened? Now we have MS. Nothing was solved.
I'll tell you what REALLY solves the problem. Instead of a state suing, why doesn't it just boycott MS products? Same with the feds. I bet the monopoly would be broken very quickly. Remember, the states are the ones perpetuating the MS monopoly by buying huge amounts of software and introducing it in schools. If California stopped buying, that alone would fracture the market enough to destroy the technology monoculture.
Most of your points just fall apart when multiple applications are accessing the same data set. In order to "never trust the engine" all the applications need to independently verify the consistency of your data.
If one application does something slightly wrong (perhaps due to a crashed webserver or whatever) if could leave inconsistent results in the database, and could cause other applications to fail, perhaps with very hard to find bugs.
For all the applications to reliably maintain the consistency of the data requires redundant code orders of magnitude more complex than simply relying on the DB engine. What if a web server crashes after it inserts some data but before it checks if it's right? What if another application accesses wrong data before the first application fixes it? All of these are non-trivial problems that take database developers a long time to get right.
I see why you came to the conclusion that you should "never trust the engine" when your engine is MySQL. However, perhaps it will save you some time if you use a database that you can trust (like, pretty much any other RDBMS).
Also, some databases are more than just a place to shove/grab data. They call them "Relational Database Management Systems". Relational mathematics (which have been developed over 30 years) are used to abstract the data from the physical storage. That's important, because application A might prefer to write data in one way, and application B might prefer to read that data in another way. RDBMSs avoid client-side manipulation or processing. You can not only get whatever data you want, you can get it in whatever form you want.
The kind of databases you're talking about are record databases, or some such. Those were available many years ago, and people decided they needed something better. Relational databases took over for a reason.
We know that MySQL records queries in the binary log after it executes them. We also know that MySQL writes transactions to the binary log after they have been committed. What happens if MySQL crashes, or someone pulls the plug in the microseconds after a transaction has been committed but before it writes the transaction to the binary log?
The result is that the master will contain the results of having completed the transaction, but the slaves will never see it. Ever. The transaction may have been a simple insert, or it could have been something as dramatic as a DROP TABLE command.
There is currently no workaround for this problem. Luckily MySQL crashes are rare. Make sure the power cables are plugged in tightly!
Wow. In PostgreSQL that would be considered a major bug. The fact that MySQL does not view it that way scares me.
In PG, vacuum has never required the DB to be offline.
In the distant past, vacuum required more problematic locks. Now it doesn't interfere with concurrent transactions. It even has built-in delays so that it effectively has a lower priority (which is configurable).
PG has been a 24x7 DB for many people for quite some time. Now, however, PG does better when under very high load 24x7.
[Schools are] where technology can be most wisely spent
You could say the same thing about any resource, and it just doesn't make sense in many situations.
My High School had computers, and the only thing I ever used them for was reading/., which required installing an HTTP proxy on a remote server to bypass the filters. I found out after I graduated, they bought class sets of wirelessly-networked mini laptops for the students.
This was in a supposedly "cash-strapped" school that strictly limited numbers of copies, etc.
To me, the whole thing is rediculous. Education is not expensive. If people really want to learn, they can learn very cheaply (like in China or India). If people don't value their education, you can spend millions and they won't learn anything.
Sweeping statements advocating the allocation of all possible resources to school sound to me like regurgitating what the teachers' unions tell you.
In Brazil, where many people can't get access to a computer, sure, have a few around for the truly interested students. But most of the computers will be more productive when used for business purposes, creating high-quality jobs.
It's not programming, it's interface design. Programmers just happen to spend a lot of time doing it. If you're one of those unfortunate programmers, you would really appreciate a drag-n-drop interface. After all, for "real programming", which is pretty much math, it makes sense to write in text. However, for interface design, it makes a lot of sense to do it visually.
I don't think there's any magical political platform that all the "average" people are going to agree on. I have pretty common political views, but I disagree with enough people that I know it would never work.
The Geneva Convention applies only to enemy combatants that also follow the Geneva Convention rules. In particular, combatants must be uniformed. As long as the people being detained at Gitmo aren't U.S. citizens, it doesn't violate the law, the Constitution, or the Geneva Convention.
I mean, whatever, how much food does Denny's throw away every day anyway?
Well, that's just stupid. Tell that to the waitress who served them and didn't get a dime when she could have gotten a real tip from non-criminals. Probably an overworked old lady who needed the money, too.
One thing that people should realize is that when an anonymous big corporation gets hurt, it's not really a corporation that's getting hurt. It's the stockholders. It's everyone that has a 401(k) or a mutual fund or any other diversified investment. These aren't rich day-traders, these are old people living off this money, or young people trying to save a few bucks for when they can't work any more.
What this guy did was every bit as noble as if he tackled a purse-snatcher. No, it's not his money, but it's someone's money, and that's a real, normal person just like you. He won't get a direct benefit, but he contributed to society.
Maybe I'm a little cynical, but if the indie artists really wanted people to share files, they would make them public domain or GPL(or GFDL or whatever it's called) or something. That would not only increase the distribution, it would be good PR, and it would lend legitimacy to online file sharing.
However, I get the idea that many of them only want to use file sharing when it suits them, and enforce the copyright later when it suits them. When people want to start buying CDs they will want a piece of the action.
Corporations are much less dangerous than the government. Only trade the former for the latter in extreme circumstances.
A corporation can do what to you? Basically, the two worst things that I can think of are: (1) Not sell you something (2) Not hire you
The government can do everything up to and including execute you.
If you have some drugs around in the privacy of your own home, the police can bust down your door, haul you to jail, get a rubber-stamp conviction from a judge, take everything you own, and sell it at auction. And that happens on a pretty regular basis. Now that is something to REALLY worry about. The parent was right: reduce the size of government, and worry about the corporate problems later, or worry about them at the state level.
Interesting. My MySQL knowledge leaves something to be desired. I looked briefly in the documentation and most of what I found was master-slave replication. Could you point me to a master-master doc?
PgPool is fairly primitive, it's just query based replication. So your application definitly needs to account for that and it isn't perfect for all situations. I mentioned it because it's the only master-master replication software that I've used (and I haven't used it except to mess around). It isn't nearly as refined as Slony-I (which is a great system), but Slony-I is master-slave.
What we're really waiting for is Slony-II, which is supposed to be a good master-master solution, but it's under development and probably won't be in widespread use for a year or so.
policy is supposed to be that guilty men go free provided no innocent person ever gets convicted
which is different from:
Better that ten guilty persons escape than that one innocent suffer.
There is a HUGE difference between those two. The former (what you said) does not even acknowledge that there is a tradeoff. The latter does, and it suggests a balance. The latter seems fine with me, but the former does not.
Yes, you do sound bitter, but it doesn't change the fact that 50% of cases of Type I diabetes are in people above the age of 18.
And, assuming that's true, that does not imply that the 50% to which you're referring did not suffer before they were 18. It merely implies that they were not diagnosed until they were 18. Which more clearly suggests the reasoning behind my rant.
I actually did not follow the case that closely... I have reasonable doubts
Hmm... that's just dumb. Maybe if you did follow the trial as closely as the jurors you would have less doubt??
The standard is "beyond a reasonable doubt". You imply that there is a reasonable doubt, which implies that there is a reasonable story.
Please produce a reasonable story that is consistent with the facts and eyewitness testimony that does not result in the guilt of Scott Peterson. For example, finish the following sentences:
(1) I am Scott Peterson, and I told my mistress that my wife was dead, even though she was still alive, because...
(2) I am Scott Peterson, and I rented a car, even though I had my own, and I drove it 80 miles away and sat looking upon the site which would, in the future, be the place Laci's body and the body of her infant were found, because...
(3) I am Scott Peterson, and I asked the police for grief counseling while there was every reason to think that Laci was still alive because...
You can see how this could start to make that transistion from the realm of "reasonable" to "unreasonable".
And wrong convictions WILL happen in any justice system, even a good one. I don't buy the idea that a wrongly convicted person is a sign of a bad justice system, I don't buy it for a second (keep in mind here that "wrongly convicted" != "innocent"). Trials are accurate the vast majority of the time. The fact that we even go back and exonerate convicts is a sign that the system is in good shape. Your standards for conviction are simply too high. If you go from "beyond a reasonable doubt" to "beyond an unreasonable doubt", you impose huge costs and you let too many guilty people go (which could have the compounding effect of not deterring enough crime). Don't get too morally superior letting guilty people walk the streets.
Now, if you want to argue that the death penalty is bad, that's a separate issue that reasonable people can disagree about.
Well, the prosecution had to go to all the trouble of proving all those facts through investigation and witnesses.
Regardless of what Scott Peterson said at trial, those facts narrowed it down to either guilt or the least plausible series of events that I've heard of.
He was innocent until proven guilty. What would it take to convince you that he's guilty if not all of the evidence that the prosecution presented? Now you're asking for him to be proven beyond all unreasonable doubt. It's just getting rediculous. The justice system gave him every presumption of innocence, including letting him hang around the Mexican border for a while with tools and $10k in cash.
A bunch of people sitting around their TV I'm sure proclaimed his guilt when there was still reasonable doubt. I heard one person say he was guilty of her murder the day Laci went missing (I used to live in Modesto). But the people that actually put him away certainly gave him every presumption of innocence until the facts were overwhelming.
They didn't hear the defense attorney at all. Jurors have to sit through all the defense stuff before they can convict anyone.
What about the other jurors? You have to be guilty beyond a reasonable doubt in the minds of 12 people at once.
It's one thing to be disgusted with the guy and say you'll lock him away, it's quite another to sit through speeches about how "Yes, he's a bad guy. He cheated. But he didn't hurt anyone." and having someone explain to you that your conviction will likely result in execution.
Slashdot Mirror
Except that's illegal fot a monopoly.
I doubt that matters. It's not really law anyway, because it's open to wildly varying interpretations. Law is objective, anti-trust "law" is subjective.
Really, who would sue? Novell couldn't last that long in court without cashflow. However long they could last would probably just result in a settlement much less generous than that with Sun & MS.
So let's say a state or the feds sue. MS would just drag it out, and since there is no smoking gun, nothing will really happen. What if the states win? What do they do? Anything the states do against MS would just be hitting the stockholders. And by stockholders, I mean the retired people living on diversified investment income. The execs have already made the money from the "anticompetitive" behavior, probably much of it in terms of huge bonuses.
Anti trust law is a mess. It doesn't really accomplish anything. After all, they went after IBM, and what happened? Now we have MS. Nothing was solved.
I'll tell you what REALLY solves the problem. Instead of a state suing, why doesn't it just boycott MS products? Same with the feds. I bet the monopoly would be broken very quickly. Remember, the states are the ones perpetuating the MS monopoly by buying huge amounts of software and introducing it in schools. If California stopped buying, that alone would fracture the market enough to destroy the technology monoculture.
Most of your points just fall apart when multiple applications are accessing the same data set. In order to "never trust the engine" all the applications need to independently verify the consistency of your data.
If one application does something slightly wrong (perhaps due to a crashed webserver or whatever) if could leave inconsistent results in the database, and could cause other applications to fail, perhaps with very hard to find bugs.
For all the applications to reliably maintain the consistency of the data requires redundant code orders of magnitude more complex than simply relying on the DB engine. What if a web server crashes after it inserts some data but before it checks if it's right? What if another application accesses wrong data before the first application fixes it? All of these are non-trivial problems that take database developers a long time to get right.
I see why you came to the conclusion that you should "never trust the engine" when your engine is MySQL. However, perhaps it will save you some time if you use a database that you can trust (like, pretty much any other RDBMS).
Also, some databases are more than just a place to shove/grab data. They call them "Relational Database Management Systems". Relational mathematics (which have been developed over 30 years) are used to abstract the data from the physical storage. That's important, because application A might prefer to write data in one way, and application B might prefer to read that data in another way. RDBMSs avoid client-side manipulation or processing. You can not only get whatever data you want, you can get it in whatever form you want.
The kind of databases you're talking about are record databases, or some such. Those were available many years ago, and people decided they needed something better. Relational databases took over for a reason.
As long as you turn off write caching, you're good.
Wow. In PostgreSQL that would be considered a major bug. The fact that MySQL does not view it that way scares me.
In PG, vacuum has never required the DB to be offline.
In the distant past, vacuum required more problematic locks. Now it doesn't interfere with concurrent transactions. It even has built-in delays so that it effectively has a lower priority (which is configurable).
PG has been a 24x7 DB for many people for quite some time. Now, however, PG does better when under very high load 24x7.
[Schools are] where technology can be most wisely spent
/., which required installing an HTTP proxy on a remote server to bypass the filters. I found out after I graduated, they bought class sets of wirelessly-networked mini laptops for the students.
You could say the same thing about any resource, and it just doesn't make sense in many situations.
My High School had computers, and the only thing I ever used them for was reading
This was in a supposedly "cash-strapped" school that strictly limited numbers of copies, etc.
To me, the whole thing is rediculous. Education is not expensive. If people really want to learn, they can learn very cheaply (like in China or India). If people don't value their education, you can spend millions and they won't learn anything.
Sweeping statements advocating the allocation of all possible resources to school sound to me like regurgitating what the teachers' unions tell you.
In Brazil, where many people can't get access to a computer, sure, have a few around for the truly interested students. But most of the computers will be more productive when used for business purposes, creating high-quality jobs.
Actually it can. The trade deficit is not a very large part of GDP. I'm not saying it should, but it can, so long as the GDP stays high.
It's not programming, it's interface design. Programmers just happen to spend a lot of time doing it. If you're one of those unfortunate programmers, you would really appreciate a drag-n-drop interface. After all, for "real programming", which is pretty much math, it makes sense to write in text. However, for interface design, it makes a lot of sense to do it visually.
He was advocating responsible gun ownership, not crime. There is evidence that gun ownership reduces violent crime.
I don't think there's any magical political platform that all the "average" people are going to agree on. I have pretty common political views, but I disagree with enough people that I know it would never work.
The Geneva Convention applies only to enemy combatants that also follow the Geneva Convention rules. In particular, combatants must be uniformed. As long as the people being detained at Gitmo aren't U.S. citizens, it doesn't violate the law, the Constitution, or the Geneva Convention.
I mean, whatever, how much food does Denny's throw away every day anyway?
Well, that's just stupid. Tell that to the waitress who served them and didn't get a dime when she could have gotten a real tip from non-criminals. Probably an overworked old lady who needed the money, too.
One thing that people should realize is that when an anonymous big corporation gets hurt, it's not really a corporation that's getting hurt. It's the stockholders. It's everyone that has a 401(k) or a mutual fund or any other diversified investment. These aren't rich day-traders, these are old people living off this money, or young people trying to save a few bucks for when they can't work any more.
What this guy did was every bit as noble as if he tackled a purse-snatcher. No, it's not his money, but it's someone's money, and that's a real, normal person just like you. He won't get a direct benefit, but he contributed to society.
Maybe I'm a little cynical, but if the indie artists really wanted people to share files, they would make them public domain or GPL(or GFDL or whatever it's called) or something. That would not only increase the distribution, it would be good PR, and it would lend legitimacy to online file sharing.
However, I get the idea that many of them only want to use file sharing when it suits them, and enforce the copyright later when it suits them. When people want to start buying CDs they will want a piece of the action.
Are you sure about that? When deciding whether something is "fair use" that's one of the things taken into account.
Corporations are much less dangerous than the government. Only trade the former for the latter in extreme circumstances.
A corporation can do what to you? Basically, the two worst things that I can think of are:
(1) Not sell you something
(2) Not hire you
The government can do everything up to and including execute you.
If you have some drugs around in the privacy of your own home, the police can bust down your door, haul you to jail, get a rubber-stamp conviction from a judge, take everything you own, and sell it at auction. And that happens on a pretty regular basis. Now that is something to REALLY worry about. The parent was right: reduce the size of government, and worry about the corporate problems later, or worry about them at the state level.
Interesting. My MySQL knowledge leaves something to be desired. I looked briefly in the documentation and most of what I found was master-slave replication. Could you point me to a master-master doc?
PgPool is fairly primitive, it's just query based replication. So your application definitly needs to account for that and it isn't perfect for all situations. I mentioned it because it's the only master-master replication software that I've used (and I haven't used it except to mess around). It isn't nearly as refined as Slony-I (which is a great system), but Slony-I is master-slave.
What we're really waiting for is Slony-II, which is supposed to be a good master-master solution, but it's under development and probably won't be in widespread use for a year or so.
You can reach me at jdavis at empires d0t 0rg.
A DFA can parse any string in time
proportional to the length of the string.
Not only that, they require only a finite amount of memory to parse an infinite string.
You siad:
policy is supposed to be that guilty men go free provided no innocent person ever gets convicted
which is different from:
Better that ten guilty persons escape than that one innocent suffer.
There is a HUGE difference between those two. The former (what you said) does not even acknowledge that there is a tradeoff. The latter does, and it suggests a balance. The latter seems fine with me, but the former does not.
policy is supposed to be that guilty men go free provided no innocent person ever gets convicted
No it's not. According to whom? I don't agree with that policy at all.
A guilty plea is a bargain with the DA to get a reduced sentence if you save the court's time by admitting guilt.
I don't see anything weird about that. He doesn't have to testify against himself.
Yes, you do sound bitter, but it doesn't change the fact that 50% of cases of Type I diabetes are in people above the age of 18.
And, assuming that's true, that does not imply that the 50% to which you're referring did not suffer before they were 18. It merely implies that they were not diagnosed until they were 18. Which more clearly suggests the reasoning behind my rant.
Do you really expect me to:
(a) notice; and
(b) believe that that is his real email address; and
(c) assume that it summarily discredits every word he writes?
Personally, I didn't make it past (a), so the compiler's optimizer in my brain never bothered to check (b) or (c).
I actually did not follow the case that closely ... I have reasonable doubts
...
...
...
Hmm... that's just dumb. Maybe if you did follow the trial as closely as the jurors you would have less doubt??
The standard is "beyond a reasonable doubt". You imply that there is a reasonable doubt, which implies that there is a reasonable story.
Please produce a reasonable story that is consistent with the facts and eyewitness testimony that does not result in the guilt of Scott Peterson. For example, finish the following sentences:
(1) I am Scott Peterson, and I told my mistress that my wife was dead, even though she was still alive, because
(2) I am Scott Peterson, and I rented a car, even though I had my own, and I drove it 80 miles away and sat looking upon the site which would, in the future, be the place Laci's body and the body of her infant were found, because
(3) I am Scott Peterson, and I asked the police for grief counseling while there was every reason to think that Laci was still alive because
You can see how this could start to make that transistion from the realm of "reasonable" to "unreasonable".
And wrong convictions WILL happen in any justice system, even a good one. I don't buy the idea that a wrongly convicted person is a sign of a bad justice system, I don't buy it for a second (keep in mind here that "wrongly convicted" != "innocent"). Trials are accurate the vast majority of the time. The fact that we even go back and exonerate convicts is a sign that the system is in good shape. Your standards for conviction are simply too high. If you go from "beyond a reasonable doubt" to "beyond an unreasonable doubt", you impose huge costs and you let too many guilty people go (which could have the compounding effect of not deterring enough crime). Don't get too morally superior letting guilty people walk the streets.
Now, if you want to argue that the death penalty is bad, that's a separate issue that reasonable people can disagree about.
Well, the prosecution had to go to all the trouble of proving all those facts through investigation and witnesses.
Regardless of what Scott Peterson said at trial, those facts narrowed it down to either guilt or the least plausible series of events that I've heard of.
He was innocent until proven guilty. What would it take to convince you that he's guilty if not all of the evidence that the prosecution presented? Now you're asking for him to be proven beyond all unreasonable doubt. It's just getting rediculous. The justice system gave him every presumption of innocence, including letting him hang around the Mexican border for a while with tools and $10k in cash.
A bunch of people sitting around their TV I'm sure proclaimed his guilt when there was still reasonable doubt. I heard one person say he was guilty of her murder the day Laci went missing (I used to live in Modesto). But the people that actually put him away certainly gave him every presumption of innocence until the facts were overwhelming.
It actually sounds somewhat reasonable. Consider:
They didn't hear the defense attorney at all. Jurors have to sit through all the defense stuff before they can convict anyone.
What about the other jurors? You have to be guilty beyond a reasonable doubt in the minds of 12 people at once.
It's one thing to be disgusted with the guy and say you'll lock him away, it's quite another to sit through speeches about how "Yes, he's a bad guy. He cheated. But he didn't hurt anyone." and having someone explain to you that your conviction will likely result in execution.