Males have greater brain specialization. (In particular, right handed males have the most.) This is why savants are more likely to be male. Head injuries to males (and especially right-handed males) are more likely to cause the complete and utter loss of some function.
So you can have female savants, you can have female geniuses, you can have just as many females doing just fine in math, but the overall likelihood is that at the very top of the field, where the people are often badly broken people who specialize in math and seem oddly incapable of anything else, the ratio of males to females will be higher.
Is this a societal phenomenon rather than a genetic one? While it might be a mix of factors, you absolutely cannot argue that male brains are just like female ones. You need only look at the prevalence of autism in males vs. females to see this. (Unless you're going to argue that autism is all about rearing technique -- in which case we ought to be dressing all our children uniformly in pink.)
Every X10 failure I ever saw (and ALL my X10 devices that switched failed over time) looked like a failure of a relay. Many posts on this discussion are talking about how X10 devices fail. The fact that they use crap relays is precisely my point. Is there something inherent to X10 (which is a protocol and a standard, not a device as such) which makes X10 unrelaiable? I don't think so -- but I would have liked a response from someone who knew the electronics of both who could say that there was or was not.
I suspect that any new standard which gets manufactured by the same companies that make X10 devices and sold in the same cost-space (which the newer specs might or might not) would have the exact same problem due to (as you put it) crap relays.
If the code size attribute is measured in number of lines, I suspect that forth, which is practically an assembly language, will rank very low (near the top of the graph, if not at the very top), though it ought to be very fast (near the left). It depends so much on stack operations that I suspect its left to right ranking would depend a great deal on the processor it's running on.
I love forth. I learned it many years ago. But I've never been in a position to use it for anything, which is a shame.
I found X10 devices failed all the time, and I've seen that mentioned here. But wasn't it the relays in them which failed, not the electronics? That means that any method of controlling a lightswitch would fail, unless you put a better switching mechanism in.
Let's all take a moment and remember how insurance companies make money, then we can proceed from there.
Insurance companies make their money on investments. When the stock market crashes, insurance companies lose money. They try to price their policies such that they make exactly ZERO dollars ($0) on insurance. They want payout and pay in to be exactly the same. That keeps them competitive, and keeps the pool of money to invest large.
Insurance companies don't care how many people die. They don't care about payouts. What insurance companies care about is predictability.
They need to know how much will be coming in and how much will be going out. If too much is going out, they raise rates, cut types of policies, etc. They mitigate their risk by being INSURED. Yes, little insurance companies buy insurance on your insurance from bigger insurance companies.
So what an insurance company wants is the largest possible pool of capital. This MAY mean that they want more accidents. More payout equals more premiums and thus possibly more cash on hand to invest.
And they want predictability. So what insurance companies really don't want is change. They know what accidents were like last year and the year before. If this tech changes the numbers, and especially if it causes a sudden plunge and then a later correction, that's really annoying and makes it harder for them to do their jobs and make their money.
Unless you're freakishly tall, try Scion Xb. Huge legroom, huge headroom. I don't know about the current model -- drive one from two years back. It'll be cheaper and easier to park, anyway.
"Now, while we're on the subject of ubiquitous factual errors, let's talk about how flat- and symmetric-winged aircraft can fly without any help from the Bernoulli effect."
Heck, yeah. It's nutty and irresponsible how we pump everyone full of the Bernoulli effect with respect to flight. With low power systems, you probably need the Bernoulli effect, but the more power you have, the more we're talking about a sled/surfboard, rather than an airfoil. This is true in old Cesnas, for goodness sake, and they are tiny and light. Still, the wing generally isn't giving you quite enough lift to keep you up when you fly with the nose completely flat. You MUST have some sledding angle against the oncoming airstream to maintain altitude.
You're ABSOLUTELY right, except that they don't ask about the server, do they? They TELL you that the server is down.
On the other hand, the amateurs who actually do anything with my servers blame EVERYTHING on the network. They want to VLAN everything. Graphs of tiny network usage versus their colossal CPU or local disk IO stats mean nothing. It is the network. They need their own network, then everything will be fine. And they know so much more than me.
Thank goodness no one remembers 5.25" disks anymore, since they used to see 3.5" disks as rigid, and would tell me they stored things on the "hard disk" when they'd put them on 3.5" floppy.
Now ALL the floppies are gone, so I just never ask people where they store anything. If I did, I suspect they'd tell me they put it in "the CPU."
But as someone who used to work for a University, let me add this:
Let's say you decide you need a masters, or at least want one. Be VERY careful about where you go to get it. Many departments in many schools see a master's program as a way to generate revenue. That's very different from the way they see undergrad and doctoral education. Those are what they REALLY do. Master's candidates come in to piggyback on the two core reasons for the department. That means that many master's program's are just a repackaging of the undergraduate classes at a high price.
That's really not going to help you, except to hand you a piece of paper. Other posts here have made a good point: You want to do a master's if you are fundamentally changing your approach. A good master's program should not be on how to program. It should be on how to lead programming teams. It should not be on how to read requirements. It should be on how to move into a company, establish the needs and goal of a company, come up with specs, and code the solution. It should feature engagements with local companies. It should change what you are and really teach you something.
Many (perhaps most) programs have no desire or intent to do that. They just want to move you through the existing resources they have at no (or very little) increase in cost to themselves. Do NOT wind up in such a program. Talk to a lot of people. Do not assume that a good name of an institution means they have a good program.
But then the manager the guy would have to work for would want to interview the OTHER guy, too.
I've interviewed too many people with a master's and no experience who couldn't write a fizzbuzz. But then, I've interviewed people with a master's AND experience and couldn't write a fizzbuzz.
A master's may not be a badge of disgrace (went back to school because he/she wasn't doing well in the workplace and thought "I know! I need more school!") but it's not a reliable mark of excellence either.
The "hobbits" represent a divergence from established models of Human evolutionary history. There are huge egos and invested in keeping things as they are. Conversely, there are careers to be made by getting people to believe something new.
Because of this, new people are going to overstate the likelihood that these little people represent a missing line, and established people will come up with all kinds of tenuous arguments saying that they are NOT.
You can't believe either side. Let them duke this out for a few years before coming to any conclusions.
The overall effect it has for reducing emissions is relatively small, and seems to pale in comparison to this. If this could replace the current carbon-black process entirely, that would be a pleasant outcome.
However, the article doesn't talk about how much it makes, how much we need, etc.
And a full-screen terminal (3270, etc.) is really just 25 punch cards. You press "Enter" and they get submitted. Your batch processes and the system returns you 25 punch cards which your smart 3270 punch card reader/editor displays for you.
Punch cards are based on the civil-war-era dollar bill because there were already machine to count and stack dollar bills.
Punch cards were IBM's most profitable product ever until the introduction of the IBM PC.
They spend more like $300 a year on printing the paper for a $20 subscription. They already take a bath on every subscriber they need to print the paper for. That's why these are intended to save the industry -- they reduce costs. Your subscription doesn't pay for the paper, ads do. There's no reason a longtime subscriber should pay anything for such a device. If it's $200 to make (I suspect it will be more) and they give it to you and stop printing your paper, they just MADE $100 bucks.
So $50 to people who might not actually read the paper and free to those they KNOW read it? Those sound like pretty good numbers to me.
"If it's that easy I look forward to seeing the wonkavader e-reader out by 2010 and in the hands of everyone."
Oh! for the budget of these newspapers. Oh! how I'd LOVE to be running that project. Heck yeah!
As for change of experience, I wonder if there are sales numbers as to how many people had a portable MP3 player before they bought an iPod. We had them, sure, but we're the techies who love such things.
"...unlike the iPod it's going to be a slower process migrating people from print to the device."
What? Why in Heaven's name would Roger say that? If these come out at $50, come with a library of great books (all free from Gutenberg et al.), and allow you to put whatever you like on them in some open format which the FOSS community can create converters for, why wouldn't it blow the iPOD sales records out of the water?
And there's no reason for them to charge more than $50. They spend the price of a Kindle printing newspapers on every subscriber every year. They can sell it for $50 with a one-year subscription to two newspapers, or give it to anyone who has been a subscriber (showing a pattern of reading) for more than two years.
The difference between this sort of thing and the Kindle or the iPod is striking. Those were both created to sell downloads, and thus try to cripple you from doing anything other than buy from Amazon or iTunes. This proposed reader is a desperate attempt to move off of an expensive process (printing papers) and onto a cheap one.
The Kindle and the iPod are designed to wring more and more money out of the consumer. These are designed to preserve a revenue stream from an advertiser. One is designed to entrance and restrict, the other to entrance and keep entranced, whatever small cost is needed to accomplish that.
If the newspapers don't make this thing explode such that EVERYBODY has one by the end of the first year, it'll be because of gross incompetence (which I'm still betting on, unfortunately) or lack of ability to produce enough of them.
Slashdot Mirror
This is missing an important point.
Males have greater brain specialization. (In particular, right handed males have the most.) This is why savants are more likely to be male. Head injuries to males (and especially right-handed males) are more likely to cause the complete and utter loss of some function.
So you can have female savants, you can have female geniuses, you can have just as many females doing just fine in math, but the overall likelihood is that at the very top of the field, where the people are often badly broken people who specialize in math and seem oddly incapable of anything else, the ratio of males to females will be higher.
Is this a societal phenomenon rather than a genetic one? While it might be a mix of factors, you absolutely cannot argue that male brains are just like female ones. You need only look at the prevalence of autism in males vs. females to see this. (Unless you're going to argue that autism is all about rearing technique -- in which case we ought to be dressing all our children uniformly in pink.)
Every X10 failure I ever saw (and ALL my X10 devices that switched failed over time) looked like a failure of a relay. Many posts on this discussion are talking about how X10 devices fail. The fact that they use crap relays is precisely my point. Is there something inherent to X10 (which is a protocol and a standard, not a device as such) which makes X10 unrelaiable? I don't think so -- but I would have liked a response from someone who knew the electronics of both who could say that there was or was not.
I suspect that any new standard which gets manufactured by the same companies that make X10 devices and sold in the same cost-space (which the newer specs might or might not) would have the exact same problem due to (as you put it) crap relays.
I like C, too, but one chooses ones language for the task at hand.
Isn't that the 'dlang' reference?
If the code size attribute is measured in number of lines, I suspect that forth, which is practically an assembly language, will rank very low (near the top of the graph, if not at the very top), though it ought to be very fast (near the left). It depends so much on stack operations that I suspect its left to right ranking would depend a great deal on the processor it's running on.
I love forth. I learned it many years ago. But I've never been in a position to use it for anything, which is a shame.
I found X10 devices failed all the time, and I've seen that mentioned here. But wasn't it the relays in them which failed, not the electronics? That means that any method of controlling a lightswitch would fail, unless you put a better switching mechanism in.
Am I wrong about this?
Let's all take a moment and remember how insurance companies make money, then we can proceed from there.
Insurance companies make their money on investments. When the stock market crashes, insurance companies lose money. They try to price their policies such that they make exactly ZERO dollars ($0) on insurance. They want payout and pay in to be exactly the same. That keeps them competitive, and keeps the pool of money to invest large.
Insurance companies don't care how many people die. They don't care about payouts. What insurance companies care about is predictability.
They need to know how much will be coming in and how much will be going out. If too much is going out, they raise rates, cut types of policies, etc. They mitigate their risk by being INSURED. Yes, little insurance companies buy insurance on your insurance from bigger insurance companies.
So what an insurance company wants is the largest possible pool of capital. This MAY mean that they want more accidents. More payout equals more premiums and thus possibly more cash on hand to invest.
And they want predictability. So what insurance companies really don't want is change. They know what accidents were like last year and the year before. If this tech changes the numbers, and especially if it causes a sudden plunge and then a later correction, that's really annoying and makes it harder for them to do their jobs and make their money.
Read "The Penultimate Truth" by Philip K. Dick.
Unless you're freakishly tall, try Scion Xb. Huge legroom, huge headroom. I don't know about the current model -- drive one from two years back. It'll be cheaper and easier to park, anyway.
They probably have a case for slander.
"Now, while we're on the subject of ubiquitous factual errors, let's talk about how flat- and symmetric-winged aircraft can fly without any help from the Bernoulli effect."
Heck, yeah. It's nutty and irresponsible how we pump everyone full of the Bernoulli effect with respect to flight. With low power systems, you probably need the Bernoulli effect, but the more power you have, the more we're talking about a sled/surfboard, rather than an airfoil. This is true in old Cesnas, for goodness sake, and they are tiny and light. Still, the wing generally isn't giving you quite enough lift to keep you up when you fly with the nose completely flat. You MUST have some sledding angle against the oncoming airstream to maintain altitude.
You're ABSOLUTELY right, except that they don't ask about the server, do they? They TELL you that the server is down.
On the other hand, the amateurs who actually do anything with my servers blame EVERYTHING on the network. They want to VLAN everything. Graphs of tiny network usage versus their colossal CPU or local disk IO stats mean nothing. It is the network. They need their own network, then everything will be fine. And they know so much more than me.
My users all call the box "the CPU".
Thank goodness no one remembers 5.25" disks anymore, since they used to see 3.5" disks as rigid, and would tell me they stored things on the "hard disk" when they'd put them on 3.5" floppy.
Now ALL the floppies are gone, so I just never ask people where they store anything. If I did, I suspect they'd tell me they put it in "the CPU."
Doh! You're absolutely right!
Wonderful post.
But as someone who used to work for a University, let me add this:
Let's say you decide you need a masters, or at least want one. Be VERY careful about where you go to get it. Many departments in many schools see a master's program as a way to generate revenue. That's very different from the way they see undergrad and doctoral education. Those are what they REALLY do. Master's candidates come in to piggyback on the two core reasons for the department. That means that many master's program's are just a repackaging of the undergraduate classes at a high price.
That's really not going to help you, except to hand you a piece of paper. Other posts here have made a good point: You want to do a master's if you are fundamentally changing your approach. A good master's program should not be on how to program. It should be on how to lead programming teams. It should not be on how to read requirements. It should be on how to move into a company, establish the needs and goal of a company, come up with specs, and code the solution. It should feature engagements with local companies. It should change what you are and really teach you something.
Many (perhaps most) programs have no desire or intent to do that. They just want to move you through the existing resources they have at no (or very little) increase in cost to themselves. Do NOT wind up in such a program. Talk to a lot of people. Do not assume that a good name of an institution means they have a good program.
It is an unfortunately reality that changing employers at a reasonable pace is the only way to get yourself on a good raise schedule.
But then the manager the guy would have to work for would want to interview the OTHER guy, too.
I've interviewed too many people with a master's and no experience who couldn't write a fizzbuzz. But then, I've interviewed people with a master's AND experience and couldn't write a fizzbuzz.
A master's may not be a badge of disgrace (went back to school because he/she wasn't doing well in the workplace and thought "I know! I need more school!") but it's not a reliable mark of excellence either.
The "hobbits" represent a divergence from established models of Human evolutionary history. There are huge egos and invested in keeping things as they are. Conversely, there are careers to be made by getting people to believe something new.
Because of this, new people are going to overstate the likelihood that these little people represent a missing line, and established people will come up with all kinds of tenuous arguments saying that they are NOT.
You can't believe either side. Let them duke this out for a few years before coming to any conclusions.
The overall effect it has for reducing emissions is relatively small, and seems to pale in comparison to this. If this could replace the current carbon-black process entirely, that would be a pleasant outcome.
However, the article doesn't talk about how much it makes, how much we need, etc.
Should have looked further down the page. Oracle does indeed have a "Linux on Power" download.
Does this mean every one of those customers ditched Oracle in the process, or is there an IBM POWER version of Oracle?
I don't see one on Oracle's standard downloads page.
And a full-screen terminal (3270, etc.) is really just 25 punch cards. You press "Enter" and they get submitted. Your batch processes and the system returns you 25 punch cards which your smart 3270 punch card reader/editor displays for you.
Punch cards are based on the civil-war-era dollar bill because there were already machine to count and stack dollar bills.
Punch cards were IBM's most profitable product ever until the introduction of the IBM PC.
They spend more like $300 a year on printing the paper for a $20 subscription. They already take a bath on every subscriber they need to print the paper for. That's why these are intended to save the industry -- they reduce costs. Your subscription doesn't pay for the paper, ads do. There's no reason a longtime subscriber should pay anything for such a device. If it's $200 to make (I suspect it will be more) and they give it to you and stop printing your paper, they just MADE $100 bucks.
So $50 to people who might not actually read the paper and free to those they KNOW read it? Those sound like pretty good numbers to me.
"If it's that easy I look forward to seeing the wonkavader e-reader out by 2010 and in the hands of everyone."
Oh! for the budget of these newspapers. Oh! how I'd LOVE to be running that project. Heck yeah!
As for change of experience, I wonder if there are sales numbers as to how many people had a portable MP3 player before they bought an iPod. We had them, sure, but we're the techies who love such things.
"...unlike the iPod it's going to be a slower process migrating people from print to the device."
What? Why in Heaven's name would Roger say that? If these come out at $50, come with a library of great books (all free from Gutenberg et al.), and allow you to put whatever you like on them in some open format which the FOSS community can create converters for, why wouldn't it blow the iPOD sales records out of the water?
And there's no reason for them to charge more than $50. They spend the price of a Kindle printing newspapers on every subscriber every year. They can sell it for $50 with a one-year subscription to two newspapers, or give it to anyone who has been a subscriber (showing a pattern of reading) for more than two years.
The difference between this sort of thing and the Kindle or the iPod is striking. Those were both created to sell downloads, and thus try to cripple you from doing anything other than buy from Amazon or iTunes. This proposed reader is a desperate attempt to move off of an expensive process (printing papers) and onto a cheap one.
The Kindle and the iPod are designed to wring more and more money out of the consumer. These are designed to preserve a revenue stream from an advertiser. One is designed to entrance and restrict, the other to entrance and keep entranced, whatever small cost is needed to accomplish that.
If the newspapers don't make this thing explode such that EVERYBODY has one by the end of the first year, it'll be because of gross incompetence (which I'm still betting on, unfortunately) or lack of ability to produce enough of them.