I have very, very mixed feelings about package managers. Although I like that I can just type 'apt-get install whatever' or equivalent, they do seem to exist primarily to fill the void left by the lack of standardized, binary installers for Linux. Maybe, what with the proliferation of distributions, this is unavoidable. But there's something strange when Linux is the OS that requires giant centralized databases.
Part of the reason for this might be that the standard Linux directory structures are a mess. This is one place where Apple really does it a lot better: Executables come in nice little packages that contain their own filesystems. Much better than spreading various crap around/etc,/var,/bin, and god-knows-where-else. Even Microsoft, I have to admit, didn't completely screw up with their 'Program Files/Program Name' structure. Apart from the general pollution that occurs in system32 [and the train wreck that is the registry (has it ever occurred to anyone that '/etc' is not conceptually very different from the registry? It's a single place, separate from program binaries and data files, to store all configuration.)], this was on the right track.
This is a bullshit article. Unfortunately, that has become the norm for Popular Mechanics.
The Intel 486 is hardly some arcane CPU that's so old that nobody knows how to program it. Anybody who can write assembly for modern PCs can write assembly for the 486. And anybody who wants to write in a higher-level language can -- because all the 486 development tools are still easily available.
If you read the article, you'll find that it presents no evidence whatsoever for its assertion that the Hubble's use of a 486 makes it harder to repair. In fact, it reads more like, "The Hubble has a 486, and damn that seems outdated to me! Maybe that's why it's so hard to fix!" Really, that's about the level of the 'logical' argument that you'll find in the article.
This is a very important point. We overuse the word pedophile -- particularly in the US. It is, according to psychologists, the attraction to prepubescent children.
The average age of puberty is 12.16 years in African-Americans and 12.88 years in whites (source). Of course, what we are really more interested in knowing is when puberty ends. This is somewhat less well defined, but I'd say that 16 is a reasonable estimate.
I don't think it is any coincidence, therefore, that this is around the age at which coming-of-age ceremonies are held in many cultures. In particular, this is when we have the "Sweet Sixteen" birthday party in the United States. This I find particularly amusing, because, although we tend not to think about this explicitly, it is in essence a debutante ball -- a ceremony advertising sexual availability.
So the modern "age of majority" approach -- 18 in the US -- seems out-of-sync with what is actually happening biologically. I think that this is a place where European countries tend to get this right, where the "sexual age of majority" is decoupled from the "age of majority for other legal purposes." In the U.S. this has been somewhat ameliorated by "Romeo-and-Juliet" laws, but these are not entirely adequate, because they only protect very young men -- in their teens, generally -- from prosecution.
The reason I bring this up, is that I suspect that psychologically-healthy people get unfairly hit by laws (e.g., statutory rape) really intended for pedophiles.
All this said, if you're downloading pictures of 8-year-old girls, do not find moral justification in my post. That is undoubtedly pedophilia, and I am not supporting you. Get help.
It is true: The street is wet when it rains. However, if the street is wet, it does not have rain.
True. But I think the point is that the probability that it rained given that the street is wet is greater than the unconditional probability of rain. It's just Bayes' rule.
If you want to show causation, then you must have a model and you must subject it to experiment
In the case of this study, this appears to be what was done. It was a randomized study with a control group and an experimental group. They kept everything else the same, and changed one variable. What is "wrong" with what they have done?
Now, I agree that there is a lot of junk science out there, and a large part of this is the result of confirmation bias (and more generally, a bias towards positive rather than negative results; it's better for one's career).
But from the little I can tell about this study from the pop article, I'm not sure what there is to complain about here.
the actual voting public for most decisions would diminish to those who care about the issue.
This might be a good idea, but it is not foolproof, because this is pretty much how lobbying works today: Organizations that care deeply about an issue -- whether it be corn ethanol, bridges in Alaska, representation of a particular ethnic group, or something else -- are the ones who dedicate resources (in the case of online voting, this would be time) to making sure that they get their way. The general public does not. The basic issue is the problem of diffuse costs and concentrated benefits. That said, it is possible that online voting could help this situation, by decreasing the cost of involvement.
A different line of argument might question your assertion that web users are smarter. Perhaps you think this because you read the comments on Slashdot instead of on Youtube?;-)
Please see my followup post if you're interested. It quotes an email that explains both the regulations and how schools can continue to offer convenient computing services on campus while remaining compliant with them. Perhaps Stanford simply didn't bother. [On the other hand, I also just learned that East Palo Alto, right next to Stanford, has a high violent crime rate (highest in the nation in '92; it has dropped since then, but still isn't great), so if Stanford seemed to be in lockdown perhaps that explains it.]
Actually, I found the email buried in old backups (grep, how I love thee).
Dear Thayer Community,
I wanted to give you advance warning of changes in the works that will
affect how you access Dartmouth's network. Central Computing Services
(PKCS) has assured me they will make sure the changes are well announced
beforehand, and this communication is an early start on this. The
bottom line is that access to Dartmouth's "regular" wireless (and
eventually wired) network will require user authentication.
The Communications Assistance for Law Enforcement Act (CALEA), enacted in 1994, required telecommunications providers to cooperate with law enforcement agencies in wiretapping telephone calls. Online services such as e-mail were specifically exempted at this time. In March 2004 CALEA was extended to include internet services. This was driven in part by the advent of voice-over-ip (VoIP) technologies, but "private networks" were specifically exempted from this change. Since this time, Dartmouth, along with many other education institutions, has been examining what it means to be a private network. Our current network is not considered private because we allow anyone to use it, regardless of whether they are affiliated with the college in any way. However, our current network would also not allow us to comply with wiretap requests received under CALEA. After reviewing costs and privacy concerns, we decided to change our network to be a private network, while at the same time retaining an unauthenticated access to the Internet. Over the summer we will be implementing this project. We will be contracting with an outside ISP to provide this unauthenticated access. Once completed, anyone who wants to use the wired network will need to be a member of the Dartmouth community, and will need to verify their identity before they can use the network. The wireless network will present options for accessing the Internet without authenticating, or accessing Dartmouth resources by authenticating. The unauthenticated option will be available to members of the local community, visitors, guests, etc. as well as any member of the Dartmouth community who wants to connect to the Internet without authenticating themselves. We are currently testing several different options, so additional information about how the authentication process will take place, and how it will affect individual users will be available as final determinations about which method to use are made.
If you have questions or concerns, please let me know.
[Sysadmin's name]
So it sounds like they're doing the absolute best thing here that they can, given the rules: They're maintaining private network status for students, but assuming ISP status for everyone else (and continuing to give out free wifi). But I expect that such altruistic behavior is unusual for sysadmins, who would avoid the not-strictly-necessary trouble of these kinds of arrangements. (E.g., this is certainly not the way things are done at Georgia Tech. Their I.T. "services" are positively Dilbert-esque.)
I'm not positive on this, but I think that part of the reason for this is laws that basically require it. AFAIK, either the college network is classified as an "internal network" (I'm not sure what the real legalese is; I'm paraphrasing), in which case it needs to store privacy-invasive and impractically-large logs of user activity, or it is classified as an ISP, in which case it avoids these issues (and associated liability) but is required to know who is on each IP, which basically necessitates restricted access and obnoxious login pages.
I say this because I did my undergrad at a school that used to keep its wifi completely open and unencrypted (Want security? Go through a VPN.) which was in fact quite wonderful. (This worked, I suppose, because it was in an idyllic little New England town, where the locals weren't a problem.) But after I left, I continued to get a few emails from various services on campus, and one was to the effect of my previous paragraph (i.e., that they were changing wifi access to meet new federal regs that they really didn't want to bother with but had to). So if I were to go back now, I get the impression that I'd be faced with login screens and such.
Say you're in a car, braking, and assume constant deceleration -- call this constant value a. Then velocity decreases linearly to zero, at which point (say this happens at time t = t_s) the acceleration discontinuously changes from a to zero. Notice that the acceleration is bounded below by 0 and above by a; it is always finite -- so all inertial forces here are bounded.
What is infinite is the jerk, which is the time-derivative of acceleration. In this example, the jerk is zero for all t not equal to t_s, and infinite at t_s. But of course this is imprecise. More precisely, the jerk can be represented by a Dirac delta function.
Hope that helps. (*Wonders is "infinite jerk" puns will follow.*)
And I should also point out that this is all a brand new phenomenon. It's been less than 20 years that the stock market has gone so grotesquely out-of-whack, throwing us into several bubble and burst cycles. See: http://www.downside.com/charts/sp500asmall.gif [downside.com]
A friend of mine noticed this when he needed to do a project for an engineering course in applications of the Fourier and related transforms. He did some very simple work with the short time Fourier transform, and saw that the high frequency content of prices in more recent periods (last 5-10 years) is much higher than that of earlier periods. Presented graphically, it jumped right out at you!
Second, during the separation, at the very end of the video, it seemed to run into the main body. Was that intentional? it wouldn't seem so.
The video you refer to is of an earlier, failed launch. My understanding, which might be mistaken, is that the lower stage continued producing thrust after it separated, and this caused it to ram into the upper stage. If that's true, then this is in complete agreement with what the video appears to show.
There's a video of the latest (successful) launch here. I also think that this video is from this launch as well, but it's not complete.
As for water: Actually, you'll see water in some NASA launches too; it's to damp vibrations during launch.
I was watching campaign ads from previous American presidential elections here -- it starts with the Eisenhower campaign and works forward -- and I was struck by how many candidates used the same rhetoric. "Change" has been a staple campaign theme for a long, long time.
It seemed there were three major types of ads:
"You don't switch horses in midstream."
"Change!!"
"He said X, but now he says NOT(X); don't trust him."
There might also have been a fourth, "Our candidate is a nice human being!"
Here are some examples of #2, "Change," below (I've quoted the last sentences from a number of the ads at the above URL):
"Vote for new American leadership. The country needs it; the world needs it. John F. Kennedy for president."
"Jimmy Carter: A leader, for a change."
"Clinton-Gore: For people, for a change."
"[George W. Bush]: A fresh start for America."
"Change" is exactly what you can expect the opposition party to be selling in any election. The only reason Obama's campaign seems novel is that we have the collective memory and attention-span of a goldfish.
Yes, and because the rest of us TRULY enjoy working with functional alcoholics.
As someone who has seen his fair share of real alcoholism, and who can count the number of drinks he's had on his hands, I still maintain that we need to relax a little about this Facebook stuff. I do not want a world where my relationship with potential employers' is like celebrities' relationship with the National Enquirer.
Translation: Never let your guard down, especially wherever other people might see you. Always sell yourself. Never allow your humanity to be visible. To admit weakness is to invite attack. To do otherwise is to be a moron with no judgment.
I also agree wholeheartedly with your attitude re: underage drinking. The only way to make sure people grow up to have a healthy attitude about alcohol is to make sure that they hide their drinking in closets where they can be sure nobody might photograph them and post their pictures on facebook.
If you're not an obsessive-compulsive paranoid, you have no business going to college, or succeeding in life.
You can be responsible and have good judgement and still hold a can of beer at a party.
Here's the thing: This is photographic evidence of a criminal activity. High-school age kids are not old enough to drink legally (assuming we're talking about the US, of course).
Now, I say that this is an absurd law. Everybody breaks it, and everybody knows it. It's such obvious hypocrisy, because you know damn well that the lawmakers who passed the law were drinking well before their 21st birthdays, just like everyone else. Yet it's on the books, so these people are criminals.
(Side note: Have you ever noticed that bars are typically built with parking lots next to them? I'm sure that's just for the "designated" drivers. *snicker* )
So returning to your comment: Can you have good judgment and be a criminal?
(The sad fact is that, in our society, I'm afraid the answer is "yes." We are all criminals. We just don't know it until we piss off the people whose corporate lawyers can scour the books until they find a convenient law for us to have broken.)
I need to start a corporation. A one-man corporation. And have it assume all responsibility for my speeding tickets, in the event that I ever get any. It'll be great. I'll drive 100mph everywhere, including past schools.
Slashdot Mirror
I have very, very mixed feelings about package managers. Although I like that I can just type 'apt-get install whatever' or equivalent, they do seem to exist primarily to fill the void left by the lack of standardized, binary installers for Linux. Maybe, what with the proliferation of distributions, this is unavoidable. But there's something strange when Linux is the OS that requires giant centralized databases.
Part of the reason for this might be that the standard Linux directory structures are a mess. This is one place where Apple really does it a lot better: Executables come in nice little packages that contain their own filesystems. Much better than spreading various crap around /etc, /var, /bin, and god-knows-where-else. Even Microsoft, I have to admit, didn't completely screw up with their 'Program Files/Program Name' structure. Apart from the general pollution that occurs in system32 [and the train wreck that is the registry (has it ever occurred to anyone that '/etc' is not conceptually very different from the registry? It's a single place, separate from program binaries and data files, to store all configuration.)], this was on the right track.
Modern Urbanism and the "Car Culture" have downsides.
Aren't these opposites?
This is a bullshit article. Unfortunately, that has become the norm for Popular Mechanics.
The Intel 486 is hardly some arcane CPU that's so old that nobody knows how to program it. Anybody who can write assembly for modern PCs can write assembly for the 486. And anybody who wants to write in a higher-level language can -- because all the 486 development tools are still easily available.
If you read the article, you'll find that it presents no evidence whatsoever for its assertion that the Hubble's use of a 486 makes it harder to repair. In fact, it reads more like, "The Hubble has a 486, and damn that seems outdated to me! Maybe that's why it's so hard to fix!" Really, that's about the level of the 'logical' argument that you'll find in the article.
This is a very important point. We overuse the word pedophile -- particularly in the US. It is, according to psychologists, the attraction to prepubescent children.
The average age of puberty is 12.16 years in African-Americans and 12.88 years in whites (source). Of course, what we are really more interested in knowing is when puberty ends. This is somewhat less well defined, but I'd say that 16 is a reasonable estimate.
I don't think it is any coincidence, therefore, that this is around the age at which coming-of-age ceremonies are held in many cultures. In particular, this is when we have the "Sweet Sixteen" birthday party in the United States. This I find particularly amusing, because, although we tend not to think about this explicitly, it is in essence a debutante ball -- a ceremony advertising sexual availability.
So the modern "age of majority" approach -- 18 in the US -- seems out-of-sync with what is actually happening biologically. I think that this is a place where European countries tend to get this right, where the "sexual age of majority" is decoupled from the "age of majority for other legal purposes." In the U.S. this has been somewhat ameliorated by "Romeo-and-Juliet" laws, but these are not entirely adequate, because they only protect very young men -- in their teens, generally -- from prosecution.
The reason I bring this up, is that I suspect that psychologically-healthy people get unfairly hit by laws (e.g., statutory rape) really intended for pedophiles.
All this said, if you're downloading pictures of 8-year-old girls, do not find moral justification in my post. That is undoubtedly pedophilia, and I am not supporting you. Get help.
It is true: The street is wet when it rains. However, if the street is wet, it does not have rain.
True. But I think the point is that the probability that it rained given that the street is wet is greater than the unconditional probability of rain. It's just Bayes' rule.
If you want to show causation, then you must have a model and you must subject it to experiment
In the case of this study, this appears to be what was done. It was a randomized study with a control group and an experimental group. They kept everything else the same, and changed one variable. What is "wrong" with what they have done?
Now, I agree that there is a lot of junk science out there, and a large part of this is the result of confirmation bias (and more generally, a bias towards positive rather than negative results; it's better for one's career).
But from the little I can tell about this study from the pop article, I'm not sure what there is to complain about here.
the actual voting public for most decisions would diminish to those who care about the issue.
This might be a good idea, but it is not foolproof, because this is pretty much how lobbying works today: Organizations that care deeply about an issue -- whether it be corn ethanol, bridges in Alaska, representation of a particular ethnic group, or something else -- are the ones who dedicate resources (in the case of online voting, this would be time) to making sure that they get their way. The general public does not. The basic issue is the problem of diffuse costs and concentrated benefits. That said, it is possible that online voting could help this situation, by decreasing the cost of involvement.
A different line of argument might question your assertion that web users are smarter. Perhaps you think this because you read the comments on Slashdot instead of on Youtube? ;-)
Please see my followup post if you're interested. It quotes an email that explains both the regulations and how schools can continue to offer convenient computing services on campus while remaining compliant with them. Perhaps Stanford simply didn't bother. [On the other hand, I also just learned that East Palo Alto, right next to Stanford, has a high violent crime rate (highest in the nation in '92; it has dropped since then, but still isn't great), so if Stanford seemed to be in lockdown perhaps that explains it.]
Dear Thayer Community,
I wanted to give you advance warning of changes in the works that will affect how you access Dartmouth's network. Central Computing Services (PKCS) has assured me they will make sure the changes are well announced beforehand, and this communication is an early start on this. The bottom line is that access to Dartmouth's "regular" wireless (and eventually wired) network will require user authentication.
The Communications Assistance for Law Enforcement Act (CALEA), enacted in 1994, required telecommunications providers to cooperate with law enforcement agencies in wiretapping telephone calls. Online services such as e-mail were specifically exempted at this time. In March 2004 CALEA was extended to include internet services. This was driven in part by the advent of voice-over-ip (VoIP) technologies, but "private networks" were specifically exempted from this change. Since this time, Dartmouth, along with many other education institutions, has been examining what it means to be a private network. Our current network is not considered private because we allow anyone to use it, regardless of whether they are affiliated with the college in any way. However, our current network would also not allow us to comply with wiretap requests received under CALEA. After reviewing costs and privacy concerns, we decided to change our network to be a private network, while at the same time retaining an unauthenticated access to the Internet. Over the summer we will be implementing this project. We will be contracting with an outside ISP to provide this unauthenticated access. Once completed, anyone who wants to use the wired network will need to be a member of the Dartmouth community, and will need to verify their identity before they can use the network. The wireless network will present options for accessing the Internet without authenticating, or accessing Dartmouth resources by authenticating. The unauthenticated option will be available to members of the local community, visitors, guests, etc. as well as any member of the Dartmouth community who wants to connect to the Internet without authenticating themselves. We are currently testing several different options, so additional information about how the authentication process will take place, and how it will affect individual users will be available as final determinations about which method to use are made.
If you have questions or concerns, please let me know.
[Sysadmin's name]
So it sounds like they're doing the absolute best thing here that they can, given the rules: They're maintaining private network status for students, but assuming ISP status for everyone else (and continuing to give out free wifi). But I expect that such altruistic behavior is unusual for sysadmins, who would avoid the not-strictly-necessary trouble of these kinds of arrangements. (E.g., this is certainly not the way things are done at Georgia Tech. Their I.T. "services" are positively Dilbert-esque.)
I'm not positive on this, but I think that part of the reason for this is laws that basically require it. AFAIK, either the college network is classified as an "internal network" (I'm not sure what the real legalese is; I'm paraphrasing), in which case it needs to store privacy-invasive and impractically-large logs of user activity, or it is classified as an ISP, in which case it avoids these issues (and associated liability) but is required to know who is on each IP, which basically necessitates restricted access and obnoxious login pages.
I say this because I did my undergrad at a school that used to keep its wifi completely open and unencrypted (Want security? Go through a VPN.) which was in fact quite wonderful. (This worked, I suppose, because it was in an idyllic little New England town, where the locals weren't a problem.) But after I left, I continued to get a few emails from various services on campus, and one was to the effect of my previous paragraph (i.e., that they were changing wifi access to meet new federal regs that they really didn't want to bother with but had to). So if I were to go back now, I get the impression that I'd be faced with login screens and such.
If they believe that mathematicians have taken up Derrida's agenda
That's Godel's incompleteness theorem, pretty much.
Say you're in a car, braking, and assume constant deceleration -- call this constant value a. Then velocity decreases linearly to zero, at which point (say this happens at time t = t_s) the acceleration discontinuously changes from a to zero. Notice that the acceleration is bounded below by 0 and above by a; it is always finite -- so all inertial forces here are bounded.
What is infinite is the jerk, which is the time-derivative of acceleration. In this example, the jerk is zero for all t not equal to t_s, and infinite at t_s. But of course this is imprecise. More precisely, the jerk can be represented by a Dirac delta function.
Hope that helps. (*Wonders is "infinite jerk" puns will follow.*)
Interesting. I doubt he knew that. (Obviously, I didn't.)
And I should also point out that this is all a brand new phenomenon. It's been less than 20 years that the stock market has gone so grotesquely out-of-whack, throwing us into several bubble and burst cycles. See: http://www.downside.com/charts/sp500asmall.gif [downside.com]
A friend of mine noticed this when he needed to do a project for an engineering course in applications of the Fourier and related transforms. He did some very simple work with the short time Fourier transform, and saw that the high frequency content of prices in more recent periods (last 5-10 years) is much higher than that of earlier periods. Presented graphically, it jumped right out at you!
Why are you conflating corporate taxes with personal income tax? Because that's what the discussion is about.
Second, during the separation, at the very end of the video, it seemed to run into the main body. Was that intentional? it wouldn't seem so.
The video you refer to is of an earlier, failed launch. My understanding, which might be mistaken, is that the lower stage continued producing thrust after it separated, and this caused it to ram into the upper stage. If that's true, then this is in complete agreement with what the video appears to show.
There's a video of the latest (successful) launch here. I also think that this video is from this launch as well, but it's not complete.
As for water: Actually, you'll see water in some NASA launches too; it's to damp vibrations during launch.
I find many references to MATSUSHITA brand optical drives. :-)
I was watching campaign ads from previous American presidential elections here -- it starts with the Eisenhower campaign and works forward -- and I was struck by how many candidates used the same rhetoric. "Change" has been a staple campaign theme for a long, long time.
It seemed there were three major types of ads:
There might also have been a fourth, "Our candidate is a nice human being!"
Here are some examples of #2, "Change," below (I've quoted the last sentences from a number of the ads at the above URL):
"Change" is exactly what you can expect the opposition party to be selling in any election. The only reason Obama's campaign seems novel is that we have the collective memory and attention-span of a goldfish.
Yes, and because the rest of us TRULY enjoy working with functional alcoholics.
As someone who has seen his fair share of real alcoholism, and who can count the number of drinks he's had on his hands, I still maintain that we need to relax a little about this Facebook stuff. I do not want a world where my relationship with potential employers' is like celebrities' relationship with the National Enquirer.
Translation: Never let your guard down, especially wherever other people might see you. Always sell yourself. Never allow your humanity to be visible. To admit weakness is to invite attack. To do otherwise is to be a moron with no judgment.
I also agree wholeheartedly with your attitude re: underage drinking. The only way to make sure people grow up to have a healthy attitude about alcohol is to make sure that they hide their drinking in closets where they can be sure nobody might photograph them and post their pictures on facebook.
If you're not an obsessive-compulsive paranoid, you have no business going to college, or succeeding in life.
...
Jesus, what is this world coming to?
You can be responsible and have good judgement and still hold a can of beer at a party.
Here's the thing: This is photographic evidence of a criminal activity. High-school age kids are not old enough to drink legally (assuming we're talking about the US, of course).
Now, I say that this is an absurd law. Everybody breaks it, and everybody knows it. It's such obvious hypocrisy, because you know damn well that the lawmakers who passed the law were drinking well before their 21st birthdays, just like everyone else. Yet it's on the books, so these people are criminals.
(Side note: Have you ever noticed that bars are typically built with parking lots next to them? I'm sure that's just for the "designated" drivers. *snicker* )
So returning to your comment: Can you have good judgment and be a criminal?
(The sad fact is that, in our society, I'm afraid the answer is "yes." We are all criminals. We just don't know it until we piss off the people whose corporate lawyers can scour the books until they find a convenient law for us to have broken.)
*sigh*
At least it's just a balloon, not a rocket; it doesn't actually put the crap into orbit. So it doesn't create space junk; it just wastes helium.
I hear it was actually chocolate.
I need to start a corporation. A one-man corporation. And have it assume all responsibility for my speeding tickets, in the event that I ever get any. It'll be great. I'll drive 100mph everywhere, including past schools.
My thoughts precisely.