Ah, Ok. It'll be interesting to see, when (if?) we actually do get fusion power plants, what their energy per volume production is, and what their requirements for raw materials are.
The turbines you're thinking of in California are actually quite a bit smaller.
Yeah, I remember them as being much smaller than the ones in Nebraska. The other thing that sticks in my memory is that, the few times I saw the California turbines, they were mostly idle.
Really minor nitpick: 16 MW isn't a truly impressive amount of power, but it's a lot bigger than the largest wind turbines in existance or planned.
Are you talking about individual turbines, or turbine farms? If a turbine can produce 3.6 MW, then some the farms I've seen produce more than 16MW. The last time I saw the farms in California, there were dozens. The farm I saw in Nebraska had a bunch of really huge ones. I don't think I've ever seen just a single wind turbine by itself, so I'm unclear whether you're talking about farms or single turbines.
Speaking of break-even... what's the big deal? I know what break-even means, but what's the practical significance of it? Is it just* a psychological achievement, or do things get much easier after break even?
Break-even has been "just around the corner" for the past 50 years. Assuming we hit break-even within the next few years, will it take another 50 to get 1Mw over break even, or will it progress faster than that? At this rate, we'll run out of fossil fuels long before we get any reasonably useful output.
I'm just imagining having to build something the size of the JET tokamak to produce 1Mw of surplus energy.
* I'm don't want to minimize the importance of psychological milestones. I just don't know if advancement in plasma physics is linear, or if the effort to get increasing amounts of energy out of the system decreases once we achieve break-even.
I live in California and pay $2.99/gallon these days. If I was paying $2.00/gallon, I'd save $600 a year. Whoopee. My sports car gets 23 mpg, and a "fuel-efficient" non-hybrid gets 35 mpg. Is my car worth the price delta? You bet it is.
I'll bite.
Your sports car gets 23M/G, which is 0.043 G/M, and is about the same as my Mitsubishi. I also have a hybrid that currently shows 44M/G (0.023G/M) over the last 4k miles. The average American puts from 12 to 15 thousand miles on their car per year, which is about what I do. Let's say 13,500 miles. California gas prices are around $3.00/G. We'll ignore the fact that it'll be $3.30 by the end of the week and probably $3.50 by the end of the month given the news that China is shutting down 7,000 coal mines and Indonesia is raising their oil prices. We'll just go with the $3.00 figure for now.
Ok. This means that the extra cost I paid for the hybrid technology has more than paid for itself in the past three years that I've owned the car. It means that, at this rate, I'll save 270 gallons of gas per year, which is $810/yr. Depending on who you listen to, this is anywhere from 1.3 to 2 times the amount of Bush's tax cut for the average middle-class American (me), which Bush claims has revitalized the American economy. That's a new low-mid range laptop per year. A new high-end cell phone. A month or two's worth of grocery bills. Two dinners for two at Le Bec Fin in Philadelphia.
Me, I blow it at ThinkGeek.com. But, yeah... I'll take the extra $70 per month.
Why else did we begin to achieve notable success in aerospace in the 60's, and then backslide to where we are now?
It is my personal belief that this is because we used to measure success by more than just a single metric in earlier decades. We used to measure success by how much stature we (as countries, companies, or cities) gained over others through our achievements. We used to need to measure success by military superiority. We used to measure success by sheer pride.
I don't think that in the past few decades we've measured anything significantly by any other metric than profit. Nobody competes with the US military, technologically, and those countries that could compete don't need to. Ironically, the only country that seems to be doing any innovation because of national pride seems to be Iran, and they're just re-inventing stuff the West has had for decades.
I'm generalizing, but I really do think that most govornments in the developed countries are pretty well controlled by corporations, so if it doesn't turn an immediate profit, it doesn't get researched. Or, rather, it gets researched but is funded at such low levels that it takes decades to produce.
...econo box from 1986 that kills the Prius... Geo metro 50Mpg average, when I drive all highway I get 58mpg
Your Geo "kills" the Prius with 58MPG? The EPA rates the Prius at 60MPG on the freeway. Heck, even our three year old Honda Civic hybrid has an average 44MPG over the past 55K miles. And we can fit five people, comfortably, in our Civic.
There's no such thing as a 1986 Geo Metro. They first started producing them in Japan in 1988, and they didn't start selling them in the US until 1989. They haven't made a Metro that got better than 40MPG since 1995, anyway. With only 50HP, the earlier models were woefully underpowered.
The Prius has five times your Geo's horsepower, gets about the same gas mileage, and is a comfortable ride. On top of that, the Prius emits so few greenhouse gasses that the Pennsylvania emissions tests can't measure them. The Honda Insight -- another hybrid vehicle -- is rated at 61/66MPG. I'd say hybrid technology is working out pretty well, and we don't have to drive around glorified go-karts in the bargain.
SMART CAR already availabe in canada and get's 60Mpg on it's own?
The SMART is a 2-seater diesel. You might as well have chosen a motorcycle for comparison. Diesel gives better gas mileage than regular petrol, but the emissions are more toxic, and the fuel is more expensive.
Hybrids aren't the solution, but they're a partial solution, until we come up with something better.
I prefer 'functional' languages over the 'dysfunctional' ones.
Now, that's funny.
What I _want_ is a functional language that I can use to generate real-time code for an embedded dsp. What functional languages are options for me?
Doesn't that depend on the DSP chip? I know almost nothing about DSPs, but don't you have to target your chip specifically?
There's an interesting article that talks about PPL, which might be interesting, but since I'm entirely ignorant about embedded systems in general and embedded DSPs in particular, I don't know if this helps.
Q. Q is pretty nice. It is a losely typed, interpreted Haskell with strong support for A/V control (including DSPs). The author is a professor in the Music Information department of the University of Mainz in Germany, so the audio libraries are correspondingly rich. If you like Haskell but hate Haskell's IO, Q is for you. The only downside is that it is interpreted.
No, it isn't. At least, it isn't pure. Most functional languages (Haskell is a noteable exception) bend this rule, and Lisp is no exception. Re-assignment is just so darned useful.
From Graham's ANSI Common Lisp, Chapter 2:
In all the sections before this, we kept to what is called "pure" Lisp---that is, Lisp without side-effects. A side-effect is some
change to the state of the world that happens as a consequence of evaluating an expression. When we evaluate a pure Lisp expression like (+ 1 2), there are no side-effects; it just returns a value. But when we call format, as well as returning a value, it prints something. That's one kind of side-effect.
Lisp contains a number of functions which produce side effects outside of IO; setq -- as you pointed out -- is one of them. Modern Lisp is functional -- it has all of the other characteristics of an FPL -- but it isn't pure.
Matz (Ruby's creator) acknowledges ripping off ideas from Lisp (but putting a friendlier face to it). Python is Lispy. Javascript has been called Lisp in C's clothing. These are all functional languages, or can be used functionally.
How do you define "functional language"? The key features of functional languages are that they (a) they reduce, or entirely eliminate, side effects, (b) have functions as first-class objects, (c) provide support for function currying, and (d) provide lambda expressions. Not all functional languages are strict about of these features, but I'd be wary of calling a language a "functional language" if it didn't at least number these as part of its core philosophy.
Among Ruby, Python, and Javascript, none provide currying and all are strongly oriented to using side effects; in particular, one of Ruby's strongest features is that it is truely object oriented, and OO is side effects. They all provide some support for closures, but only Javascript can claim to treat functions as first-class objects. Consequently, none of those languages are anywhere close to being functional in nature, and although you might make be able to make a case for Javascript, purists will tell you that any language that has variables (as in, able to re-assign the value of it) isn't functional.
That said, you're not the first person I've heard say that all languages are, ultimately, Lisp machines, and there's some merit to that. You can do procedural, OO, or functional style programming in almost any language, if you bend over backwards far enough; however, the term "functional programming language" does have meaning and shouldn't be applied arbitrarily just because you can write code that looks functional -- perhaps moreso than other paradigms because functional languages are often more strict about purity than other languages.
They even make sure that the speed at which the round leaves the rifle is high enough so that the round doesn't start to tumble before a given distance
This isn't true;.223 rounds do tumble and change direction in the body.
There have been a lot of posts with accurate information about the rounds used by the US military, but one thing that has been missed is that the main reason the US uses rounds with a relatively low instant fatality is that it takes three soldiers to take handle one wounded soldier. The ideal situation is to wound one enough that two of his buddies have to take time out to drag him to safety, a medic has to spend time trying to keep him alive, the opposing force has to spend time and resources evacuating him, and then he sits in a hospital wasting more time, money and resources before finally dying of his wounds. The whole goal is to make every wounded soldier as expensive as possible for the enemy.
This may be changing, as the nature of our opponents change.
I'd like to say something nasty about the current conflicts in Afghanistan and Iraq, but I don't want to politicize the thread.
"I'd be open to switching my OS if a new OS did everything that my existing OS did *and* added a bunch of new stuff that made the effort worthwhile."
...
I think that's the crux of the problem, but also the most mis-understood part. That's the part that OS zealots love to mis-understand.
Hm.
I'd switch for much less than "everything my current OS does". I have a narrow set of features that I'd like to see. Specifically, I don't use at least 90% of the software that is available for Linux, and I don't use probably 60% of the kernel features.
However, if we limit ourselves to kernel features (the OS, strictly) I'd switch if it was reasonably POSIX compliant and GCC was available for it (meaning I could compile most of the software that I use), had a fast GUI toolkit, was reliable, and had OS persistence. If I could unplug my computer at any moment and boot back up into the state just before I unplugged it, that alone would sell me. If the OS booted to a usable state in 10 seconds or less after BIOS initialization, it would be a big selling point.
It takes 40 seconds to get a login prompt in Linux and over a minute to get a KDE desktop -- the latter is, admittedly, not the kernel's fault, but is part of the larger problem. Add to this the fact tha ACPI sleep/suspend support is practically useless on all but a handful of laptops, and Linux is almost useless for people who lug their machines around with them.
I'm not blaming anybody for this, mind you. ACPI was always a shoddily implemented standard, and supporting it on the myriad implementations is daunting -- I don't envy the ACPI developers their task. And, while there are efforts to improve the init process, the sheer number of configuration permutations makes that extremely difficult as well. These are both specific cases of the more general problem that there is no solid standards body for Linux, making it exceedingly difficult for projects like KDE to support things like plug-and-play peripherals.
All of this is evident when you use a Mac. Apple strictly controls the hardware -- and the software -- and it allows them to provide a system that "Just Works"(tm). Try to do that for Linux, and you get bogged down in tweaking your software for fringe cases, and trying to track down esoteric bugs that only happen for one guy in Minosota, who's bug report consists entirely of "Hey, my machine locks up when I try to use your software. How do I fix it?".
The purpose of language is communication, and the standardization of such is to ensure against ambiguity, right? If someone's written work is devoid of some common rules of grammar and usage, does it matter if you completely and unambiguously understand what they are saying/writing?
Yes, it does matter.
First, misspellings and poor grammar are, at least, distracting. At worst, they can make the text difficult to read.
Second, -- and I'm using an IT example because that's what I'm most familiar with -- they make one look sloppy. Why would I trust a person who's exhibited obvious linguistic carelessness with any responsibility? You might as well have said, "if a program is devoid of code comments and optimization, does it matter if gives you the correct result?" It may do the job, but if I have a choice, I'm going to choose a programmer who can actually code properly.
That said, I do believe that Slashdot posts are special cases that deserve some leniency. Posting to Slashdot is less writing literature and more conversation, and conversation has always had looser rules (remember the old "a lot" rule)?
Although, with Freenet, it isn't ignorance of the law that protects you, but ignorance of the packet data.
Freenet is designed such that any given file is broken up into packets, and that it is not possible -- given a packet -- to tell what the content of the entire file is. They've implemented this design with varying degrees of success, but the point is that:
You have some cache that has some packets in it. Some of these packets comprise files you've injected into the system, and some are packets that you've "captured" as requests have gone through your node. For these other packets, you don't know what their content is, and you have no way of discovering this information [1]
When you request a file, you connect to some nodes and pass along a request for packets with t. These nodes either return the packets matching the hashes you're requesting, if they have them, or pass along the request to nodes they're connected to. In either case, when they do finally return the packets to you, you have no idea whether those packets were cached on that node, or some other node down the tree [2]
Therefore, your culpability is limited to possibly caching parts of a file which you can reasonably deny knowledge of the contents, and to having relayed parts of a file with the same reasonable deniability.
Caveat lector
It has been, literally, a couple of years since I've looked at Freenet, and I'm certain that the design and implementation has changed. However, the concepts are pretty clear, and the only challenge is in the implementation.
There is a way to decode the contents of a packet: by requesting the file which the packet belongs to, and comparing the packets you get. But the file hash the packet belongs to is not encoded with the packet itself, so this solution has an upper bound of N, for N files in the network.
There are attacks for this, such as response time, and problems raised in implementing this algorithm, such as time-to-live. Again, these have been solved with varying degree of success.
# The first time a unit is actually ordered and purposefully told to attack a rioting/rebellious crowd.
You might have something with this. US Servicemembers are men (and women) of honor. However, It'd have to escalate quite a bit before lethal force would be required.
Jezus. Does nobody remember Kent State? When the national guard fired into a crowd of student protesters? Hello? How about Abu Graib prison?
Some US service members are people of honor. I think it's gross exaggeration to imply that all, or even most, of them are. A good percentage of the men I served with in the Army were right bastards, and your own comments ("I'd find it rather funny watching the crowd on their asses in super-slime engulfed in CS gas") make me wonder what your definition of "honor" is. Gassing unarmed civilians, and laughing about it, isn't quite how I'd define "honorable".
This is the same kind of black and white reasoning that George W. Bush uses. "You're either with us, or you're with the terrorists."
No, it is more like saying "You either approve of the 13th amendment, or you're a racist." There's a difference.
Although, you're fundamentally right; MS is going neutral because they think it will be better for business, not because they've suddenly become intolerant of gays and lesbians.
Of course, this just emphasizes the total lack of any ethics in the company; no matter which side you're on, a person (or company) who changes an opinion for expediency is always pathetic. Successful, perhaps, but certainly pathetic, unethical, and undependable.
Slashdot Mirror
Ooo, look! Somebody gave the Republicans mod points!
--- SER
How did you do your calculations? I get:
Would you mind elaborating?
--- SER
Are you talking about individual turbines, or turbine farms? If a turbine can produce 3.6 MW, then some the farms I've seen produce more than 16MW. The last time I saw the farms in California, there were dozens. The farm I saw in Nebraska had a bunch of really huge ones. I don't think I've ever seen just a single wind turbine by itself, so I'm unclear whether you're talking about farms or single turbines.
--- SER
Break-even has been "just around the corner" for the past 50 years. Assuming we hit break-even within the next few years, will it take another 50 to get 1Mw over break even, or will it progress faster than that? At this rate, we'll run out of fossil fuels long before we get any reasonably useful output.
I'm just imagining having to build something the size of the JET tokamak to produce 1Mw of surplus energy.
* I'm don't want to minimize the importance of psychological milestones. I just don't know if advancement in plasma physics is linear, or if the effort to get increasing amounts of energy out of the system decreases once we achieve break-even.
--- SER
Your sports car gets 23M/G, which is 0.043 G/M, and is about the same as my Mitsubishi. I also have a hybrid that currently shows 44M/G (0.023G/M) over the last 4k miles. The average American puts from 12 to 15 thousand miles on their car per year, which is about what I do. Let's say 13,500 miles. California gas prices are around $3.00/G. We'll ignore the fact that it'll be $3.30 by the end of the week and probably $3.50 by the end of the month given the news that China is shutting down 7,000 coal mines and Indonesia is raising their oil prices. We'll just go with the $3.00 figure for now.
Ok. This means that the extra cost I paid for the hybrid technology has more than paid for itself in the past three years that I've owned the car. It means that, at this rate, I'll save 270 gallons of gas per year, which is $810/yr. Depending on who you listen to, this is anywhere from 1.3 to 2 times the amount of Bush's tax cut for the average middle-class American (me), which Bush claims has revitalized the American economy. That's a new low-mid range laptop per year. A new high-end cell phone. A month or two's worth of grocery bills. Two dinners for two at Le Bec Fin in Philadelphia.
Me, I blow it at ThinkGeek.com. But, yeah... I'll take the extra $70 per month.
It is my personal belief that this is because we used to measure success by more than just a single metric in earlier decades. We used to measure success by how much stature we (as countries, companies, or cities) gained over others through our achievements. We used to need to measure success by military superiority. We used to measure success by sheer pride.
I don't think that in the past few decades we've measured anything significantly by any other metric than profit. Nobody competes with the US military, technologically, and those countries that could compete don't need to. Ironically, the only country that seems to be doing any innovation because of national pride seems to be Iran, and they're just re-inventing stuff the West has had for decades.
I'm generalizing, but I really do think that most govornments in the developed countries are pretty well controlled by corporations, so if it doesn't turn an immediate profit, it doesn't get researched. Or, rather, it gets researched but is funded at such low levels that it takes decades to produce.
--- SER
There's no such thing as a 1986 Geo Metro. They first started producing them in Japan in 1988, and they didn't start selling them in the US until 1989. They haven't made a Metro that got better than 40MPG since 1995, anyway. With only 50HP, the earlier models were woefully underpowered.
The Prius has five times your Geo's horsepower, gets about the same gas mileage, and is a comfortable ride. On top of that, the Prius emits so few greenhouse gasses that the Pennsylvania emissions tests can't measure them. The Honda Insight -- another hybrid vehicle -- is rated at 61/66MPG. I'd say hybrid technology is working out pretty well, and we don't have to drive around glorified go-karts in the bargain.
The SMART is a 2-seater diesel. You might as well have chosen a motorcycle for comparison. Diesel gives better gas mileage than regular petrol, but the emissions are more toxic, and the fuel is more expensive.Hybrids aren't the solution, but they're a partial solution, until we come up with something better.
There's an interesting article that talks about PPL, which might be interesting, but since I'm entirely ignorant about embedded systems in general and embedded DSPs in particular, I don't know if this helps.
Two other languages that might interest you are:
--- SER
From Graham's ANSI Common Lisp, Chapter 2:
Lisp contains a number of functions which produce side effects outside of IO; setq -- as you pointed out -- is one of them. Modern Lisp is functional -- it has all of the other characteristics of an FPL -- but it isn't pure.--- SER
Among Ruby, Python, and Javascript, none provide currying and all are strongly oriented to using side effects; in particular, one of Ruby's strongest features is that it is truely object oriented, and OO is side effects. They all provide some support for closures, but only Javascript can claim to treat functions as first-class objects. Consequently, none of those languages are anywhere close to being functional in nature, and although you might make be able to make a case for Javascript, purists will tell you that any language that has variables (as in, able to re-assign the value of it) isn't functional.
That said, you're not the first person I've heard say that all languages are, ultimately, Lisp machines, and there's some merit to that. You can do procedural, OO, or functional style programming in almost any language, if you bend over backwards far enough; however, the term "functional programming language" does have meaning and shouldn't be applied arbitrarily just because you can write code that looks functional -- perhaps moreso than other paradigms because functional languages are often more strict about purity than other languages.
The exception being O'Caml, which is a mutant.
--- SER
--- SER
Go figure.
--- SER
This isn't true; .223 rounds do tumble and change direction in the body.
There have been a lot of posts with accurate information about the rounds used by the US military, but one thing that has been missed is that the main reason the US uses rounds with a relatively low instant fatality is that it takes three soldiers to take handle one wounded soldier. The ideal situation is to wound one enough that two of his buddies have to take time out to drag him to safety, a medic has to spend time trying to keep him alive, the opposing force has to spend time and resources evacuating him, and then he sits in a hospital wasting more time, money and resources before finally dying of his wounds. The whole goal is to make every wounded soldier as expensive as possible for the enemy.
This may be changing, as the nature of our opponents change.
I'd like to say something nasty about the current conflicts in Afghanistan and Iraq, but I don't want to politicize the thread.
Yeah, but your data density sucked.
--- SER
I'd switch for much less than "everything my current OS does". I have a narrow set of features that I'd like to see. Specifically, I don't use at least 90% of the software that is available for Linux, and I don't use probably 60% of the kernel features.
However, if we limit ourselves to kernel features (the OS, strictly) I'd switch if it was reasonably POSIX compliant and GCC was available for it (meaning I could compile most of the software that I use), had a fast GUI toolkit, was reliable, and had OS persistence. If I could unplug my computer at any moment and boot back up into the state just before I unplugged it, that alone would sell me. If the OS booted to a usable state in 10 seconds or less after BIOS initialization, it would be a big selling point.
It takes 40 seconds to get a login prompt in Linux and over a minute to get a KDE desktop -- the latter is, admittedly, not the kernel's fault, but is part of the larger problem. Add to this the fact tha ACPI sleep/suspend support is practically useless on all but a handful of laptops, and Linux is almost useless for people who lug their machines around with them.
I'm not blaming anybody for this, mind you. ACPI was always a shoddily implemented standard, and supporting it on the myriad implementations is daunting -- I don't envy the ACPI developers their task. And, while there are efforts to improve the init process, the sheer number of configuration permutations makes that extremely difficult as well. These are both specific cases of the more general problem that there is no solid standards body for Linux, making it exceedingly difficult for projects like KDE to support things like plug-and-play peripherals.
All of this is evident when you use a Mac. Apple strictly controls the hardware -- and the software -- and it allows them to provide a system that "Just Works"(tm). Try to do that for Linux, and you get bogged down in tweaking your software for fringe cases, and trying to track down esoteric bugs that only happen for one guy in Minosota, who's bug report consists entirely of "Hey, my machine locks up when I try to use your software. How do I fix it?".
--- SER
First, misspellings and poor grammar are, at least, distracting. At worst, they can make the text difficult to read.
Second, -- and I'm using an IT example because that's what I'm most familiar with -- they make one look sloppy. Why would I trust a person who's exhibited obvious linguistic carelessness with any responsibility? You might as well have said, "if a program is devoid of code comments and optimization, does it matter if gives you the correct result?" It may do the job, but if I have a choice, I'm going to choose a programmer who can actually code properly.
That said, I do believe that Slashdot posts are special cases that deserve some leniency. Posting to Slashdot is less writing literature and more conversation, and conversation has always had looser rules (remember the old "a lot" rule)?
--- SER
--- SER
Personally, I prefer the vertical configuration.
--- SER
Freenet is designed such that any given file is broken up into packets, and that it is not possible -- given a packet -- to tell what the content of the entire file is. They've implemented this design with varying degrees of success, but the point is that:
Caveat lector It has been, literally, a couple of years since I've looked at Freenet, and I'm certain that the design and implementation has changed. However, the concepts are pretty clear, and the only challenge is in the implementation.
--- SER
Jezus. Does nobody remember Kent State? When the national guard fired into a crowd of student protesters? Hello? How about Abu Graib prison?
Some US service members are people of honor. I think it's gross exaggeration to imply that all, or even most, of them are. A good percentage of the men I served with in the Army were right bastards, and your own comments ("I'd find it rather funny watching the crowd on their asses in super-slime engulfed in CS gas") make me wonder what your definition of "honor" is. Gassing unarmed civilians, and laughing about it, isn't quite how I'd define "honorable".
--- SER
Pretty girls are fairly common to all nations; the only difference is that the French ones know how to dress.
--- SER
No, it is more like saying "You either approve of the 13th amendment, or you're a racist." There's a difference.
Although, you're fundamentally right; MS is going neutral because they think it will be better for business, not because they've suddenly become intolerant of gays and lesbians.
Of course, this just emphasizes the total lack of any ethics in the company; no matter which side you're on, a person (or company) who changes an opinion for expediency is always pathetic. Successful, perhaps, but certainly pathetic, unethical, and undependable.
--- SER