If the reviewer wants to focus on value, it would be helpful to present each benchmark result as a scatter plot with price on one axis and performance on the other. Another set of scatter plots could present power against performance.
That would be more helpful than pretending that price and power trade-offs are the same for every reader. $87 means more to some people than to others.
You should be worried. If the MPAA/RIAA can convince consumers that it's in their best material interest to accept every legal restriction producers ask for, it will be a new economic religion that will spill over into political thinking, summed up as, "We're happy because the elites are working hard for our benefit: let no one interfere with them!"
Assertive and inquisitive consumers are crucial to the economy, and assertive and inquisitive citizens are crucial to democratic society. This must be stopped.
Oops, my comment below is mostly redundant to yours.... I disagree that the plain block is more useful in C than in C++, however. When using the "resource acquisition is initialization" (RAII) idiom, you have to pay attention to when destructors are called, since the destructor of a resource-controlling object release that resource. Although you would prefer to write small enough functions that it isn't necessary, sometimes you have to write:
{ MutexLock lock(non_threaded_library_mutex); x = non_threaded_library_foo(y); x = non_threaded_library_twiddle(x); }
This ensures that the mutex is released as quickly as possible, which decreases contention if the rest of the function body takes any significant time.
I agree with your point that while(1) can be the most elegant way to express a loop, but you left out the best and simplest way to limit the scope of a C++ variable. Here's an example (sorry, preview shows the indentation being lost, and I can't figure out how to fix it):
Creating first Creating second Destroying second Creating third Destroying third Destroying first
When adding debug or trace code into an enormous function that you don't understand, enclosing the code in { } prevents compile failures from twice-defined variables. (It won't stop you from accidentally twiddling variables from the enclosing scope, so be careful!)
The most common use is to limit the scope of objects that manage resources using RAII, such as mutex locks and database connections. Personally, I try to avoid functions large enough to need this, except in the case of mutex locks and other concurrency mechanisms, whose scope should always be as limited as clarity allows.
Sitting around night after night with the same people, hiding behind made-up identities mathematically parameterized from crude stereotypes, looking up rules in books?
Gee, I wouldn't expect to find computer geeks doing that; it's way too free-form and social.
are you listening to me now? or do you want to continue to talk about a subject matter i'm not talking about?
I'm not sure what you're talking about. You deny that psychology can legitimately speak about personality, and you deny that medicine has anything to say about human behavior. Doctors study and treat human behavior (Hans Asperger himself was a pediatrician), and you say that manic depression and schizophrenia are medical pathologies, so I don't understand how you can maintain that the medical perspective is irrelevant to personality. You have never explained your distinction between "personality" and "psychology" -- is it mind/body? Pathology/non-pathology? I'm not holding my breath to find out.
I'll repeat one last time: I agree that people are keen to pathologize anything they can sell a treatment for. We don't disagree about that. I don't agree that attacking medicine and arbitrarily declaring "personality" to be a domain distinct from "psychology" helps anything.
Denying medicine a voice merely because corporations exploit misconceptions about medicine is just silly. Medicine is just one avenue for ripping people off, and declaring psychiatry illegitimate just herds the ignorant and vulnerable toward a different set of quacks. After all, when it comes to pathologizing and "curing" behavior, medicine is still the new kid on the block compared to religion.
we have STOP the medicalization of personality
in psychology, with manic depression and schizophrenia, the language is perfectly appropriate
but with personality?
Okay, I finally get it. You think that the study of human mind and behavior is split into two separate disciplines: a "medical" branch called "psychology" (not psychiatry?) that concerns itself with pathologies and a "personality" branch for things that aren't bad and nasty. Thus making value judgements a compulsory and fundamental part of studying human behavior. No wonder you're upset.
Actually, the study of personality is usually seen as a branch of psychology, the medical branch of psychology is called psychiatry, and whether to classify Asperger's Syndrome or homosexuality as pathologies, as they were regarded by the first people to refer to them as such, is up to the individual and is not inherent in the words. Psychologists are not professionally bound to disapprove of everything they study. Even psychiatrists disagree about what does and does not constitute a pathology, and in practice, it's common to allow the patient to decide which behaviors need to be modified and which behaviors the patient is happy with.
WHY, please tell me, WHY do we need a PILL for SHYNESS???
WHY do we need to talk about personalities in this medical terminologistic way???
Now I also understand why you think these two things are connected. Indeed there is a misconception, often exploited by unscrupulous doctors and businessmen, that every medical term (or psychological term, by your language) refers to a pathology that must be treated. There's even a whole set of jokes about it (most of which are probably funnier than this one):
Doctor: Mrs. Wilson, you have something called an "aorta" that originates at your heart and branches out to several other parts of your body.
Mrs. Wilson: Oh my gosh! Is it serious?
But that's nothing more than a misconception, and a fading one at that. When my dentist refers to my "upper bicuspids" I don't conclude that he's out to cure me of them. When a bodybuilder uses the word "hypertrophy," would you assume he's opposed to having bigger muscles?
It's a comical and hopeless point of view that elevates one way of seeing the world to such a position that its mere ability to speak grants it complete authority over a topic. You seem to believe that the legitimacy of the medical or psychological perspective *must* be denied, because otherwise right and wrong becomes the domain of medicine or psychology. This is exactly the kind of thinking that leaves people at the mercy of doctors who create demand for treatment by defining more and more pathologies.
if your position is now that asperger's is not something valid to be talked about in medical terminology, that it is a normal personality type with pluses and minuses like any other, with nothing medical implied in any way whatsoever, then i thank you for backtracking
Talking about something in medical terminology doesn't implicitly deprecate it. "Medical" does not mean "pathological" or even "abnormal," as you seem to think it does.
You are worried about a future where difference is pathologized and everyone is medicated into sameness. You want "MORE tolerance and acceptance of a range of quirks."
Yet you object to an article that presents Asperger's Syndrome as a normal, liveable condition that involves advantages and disadvantages and can lead to unique contributions to society.
How is this supposed to make sense? My interpretation that you were objecting to the arbitrary discretization of a spectrum was the only sensible reading I could see, so I went with it.
You have a point, but it's just the way doctors think and communicate. They don't do math; they're not scientists who deal with fields or probability distributions. They're trained to look at a patient, apply a set of labels, search through their medical school training for rules involving those labels, and then apply those rules. They investigate continuous phenomena by dividing the spectrum into discrete chunks, each with a label, and systematically creating rules involving those labels. That's why medical diagnostic systems are such a successful application of classical AI: once the doctors handle the subjective task of applying labels, the only difference between a doctor and Prolog is that Prolog can operate faster, more reliably, and using more rules than any single doctor.
Some researchers may come up with theories about continuous phenomena, but it's alien to the way doctors think.
We aren't going to lose the next generation of kids just because they're into a bunch of stupid kid things like shaking IM windows, sending Flash animations, and having cool avatars. Honestly, do you think they're still going to be doing those things when they grow up?
Linux IM clients should integrate the features that their current users find useful and not worry that today's teenagers will be just as frivolous, shallow, and glitz-oriented when they're 25 as they are now at 15. Adults have been fretting about that prospect for thousands of years, and it hasn't happened yet.
Young people have a strong bias against old and busted-looking stuff; they'll almost always favor the newest, flashiest, slickest thing. To them, Linux instant messaging must seem simply broken.
The good news is that they'll grow up and get tired of doing all the stupid stuff that only MSN enables them to do. Shaking someone's IM window is obnoxious even by teenager standards. Exchanging files is the one thing I saw in the article that I would like to do using instant messaging. Fix that (if it doesn't work already) and you'll have everyone over 18.
You can't call it "reading" and "writing." Too many computer people think that "reading" and "writing" are fast, straightforward, linear techniques that result in a disorganized mess - which for them is probably the truth. They would never understand that putting serious mental effort into creating a coherent, useful work could be called "writing," or that a systematic, intelligent effort to understand something could be called "reading."
Parent is modded Funny; I would say Insightful. I'd love to have recommender programs that know my idiosyncrasies and nudge me out of ruts and towards things that might disturb my mindset.
...is what's on the line here, not the security of Firefox. Installing someone else's code is always insecure if they can't be trusted. I still trust Mozilla, but I'm really shocked by this.
I agree that XML relieves the burden of parsing and unparsing, and is thus superior for any configurating file that isn't just a flat list of records. In theory, that is. The available tools just don't allow XML to live up to its potential.
I'm never tempted to use XML formats when I know I'll have to edit them. Do I edit XML in a text editor, write a custom data editor backed by XML, or write a custom parser and unparser? Usually I choose the last option. I almost never use XML for files that humans see.
Where are the editing and composing tools? Where are the XML editors that aren't slow, clunky, or oriented to a few specific document formats?
The article asserts a variety of ludicrous ideas as common conceptions about OSS. It's impossible to take seriously.
I'll grant that the point about conceptual integrity may have merit. Distributed development makes conceptual integrity very hard to maintain. But how do I know that? Through commercial experience. It only applies to OSS because almost all OSS projects are distributed.
Frankly, the ideas attributed in this article to OSS people are so alien and fantastic that I doubt the author has even read any of the basic writings about open source or studied a single open source project.
Ah, I see! It's like in calculus where you solve for f'(x) = 0 to find the minima and maxima of f.
This is a perfect case study of media spin, then. Should be fun.
I think Bruce's point #5 also refers to the usefulness of code-signing in protecting against inside jobs. Since a cryptographic signature can't be faked, it can be used to hold the signer responsible for the signed content. However, as Bruce points out, signed code can let the signer off the hook by erasing its own signature. In this respect signatures on code are less useful than signatures on other kinds of content.
Software development is business, too. What do you want to bet that the WinHoldEm developers run an unreleased poker bot that has a special knack for beating WinHoldEm? It's like selling milkers to cows, then showing up with a bucket.
You don't live in a rural area. You live in the suburbs. Would the good schools and DSL would be there if it weren't for people commuting into Kansas City? Get a clue yourself.
You're right. A GENI project goal, via the article:
privacy and accountability and vary protections for individuals based on "difference and local values"
"Difference and local values" is code for enabling the Saudi Government to block DanicaRacing.com.
It's also code for enabling Bush to track and harass people who read Islamic web sites, and enabling Clinton (the current one) to make sure that "parents" have absolute power to mould their "children." Bush and Clinton (and Hatch and Kennedy and...) will first agree 100% that governments need absolute control and knowledge, and then they'll fight over what that control is used for.
In other words... theory! Information theory, compiler theory, (abstract) algebra, linear algebra, programming (non-fad-specific books such as The Science of Programming), parallel concepts, algorithms, concurrency theory, electromagnetism, etc. You should include all levels from high school to intro grad level. Include some general-reading books like Knuth's Things a Computer Scientist Rarely Talks About.
Books like that don't have to be bought and replaced often. Who cares if a compiler theory textbook is five years old, as long as the code examples aren't in Java 1.1? Leave the short-lived application-specific and language-specific stuff (except slow-changing languages like Common Lisp and historical languages like Algol) to bookstores.
Also, you should provide a way to do simple, nondangerous programming on library computers -- turtle graphics type stuff for the kiddies. You gotta make sure those future donors have fond memories:-)
Finally, you should not be disappointed that so many people here are talking about books -- you should swallow your shiny ambitions and take it to heart. It makes me sick that libraries are now pretty cool places to hang out, but I have to BRING MY OWN FRICKING BOOKS BECAUSE THE LIBRARY DOESN'T HAVE ANY. Ask yourself if anyone comes to your library for something they couldn't get at Starbucks. Your library, and libraries everywhere, may ultimately suffer the same fate as every other Starbucks competitor.
Even as someone who abandoned Perl for Python, I suspect there is a point to Perl. I'm truly impressed by the concise power of Perl one-liners and text-manipulation scripts, and I made an earnest attempt to become a Perl user.
Unfortunately, I only need scripting once every few months or weeks, which is longer than it takes me to forget all the little details that make Perl so rich and expressive. Python is much better for me: it only takes me a few minutes to start writing good code in Python, no matter how long it's been since the last time.
My failure should give some idea of who can and should use Perl:
1) People who need and use scripting every day or at least several days per week 2) People who are much smarter than me:-)
Seriously, Perl is a complex and sophisticated tool that rewards consistent commitment. You just can't half-ass it, or neglect your Perl skills and expect them to be there months later when you need them. For that, use Python.
Slashdot Mirror
If the reviewer wants to focus on value, it would be helpful to present each benchmark result as a scatter plot with price on one axis and performance on the other. Another set of scatter plots could present power against performance.
That would be more helpful than pretending that price and power trade-offs are the same for every reader. $87 means more to some people than to others.
You should be worried. If the MPAA/RIAA can convince consumers that it's in their best material interest to accept every legal restriction producers ask for, it will be a new economic religion that will spill over into political thinking, summed up as, "We're happy because the elites are working hard for our benefit: let no one interfere with them!"
Assertive and inquisitive consumers are crucial to the economy, and assertive and inquisitive citizens are crucial to democratic society. This must be stopped.
This ensures that the mutex is released as quickly as possible, which decreases contention if the rest of the function body takes any significant time.
Output:
When adding debug or trace code into an enormous function that you don't understand, enclosing the code in { } prevents compile failures from twice-defined variables. (It won't stop you from accidentally twiddling variables from the enclosing scope, so be careful!)
The most common use is to limit the scope of objects that manage resources using RAII, such as mutex locks and database connections. Personally, I try to avoid functions large enough to need this, except in the case of mutex locks and other concurrency mechanisms, whose scope should always be as limited as clarity allows.
Sitting around night after night with the same people, hiding behind made-up identities mathematically parameterized from crude stereotypes, looking up rules in books?
Gee, I wouldn't expect to find computer geeks doing that; it's way too free-form and social.
I'm not sure what you're talking about. You deny that psychology can legitimately speak about personality, and you deny that medicine has anything to say about human behavior. Doctors study and treat human behavior (Hans Asperger himself was a pediatrician), and you say that manic depression and schizophrenia are medical pathologies, so I don't understand how you can maintain that the medical perspective is irrelevant to personality. You have never explained your distinction between "personality" and "psychology" -- is it mind/body? Pathology/non-pathology? I'm not holding my breath to find out.
I'll repeat one last time: I agree that people are keen to pathologize anything they can sell a treatment for. We don't disagree about that. I don't agree that attacking medicine and arbitrarily declaring "personality" to be a domain distinct from "psychology" helps anything.
Denying medicine a voice merely because corporations exploit misconceptions about medicine is just silly. Medicine is just one avenue for ripping people off, and declaring psychiatry illegitimate just herds the ignorant and vulnerable toward a different set of quacks. After all, when it comes to pathologizing and "curing" behavior, medicine is still the new kid on the block compared to religion.
Just ask Tom Cruise.
-- and I'm out of here
in psychology, with manic depression and schizophrenia, the language is perfectly appropriate
but with personality?
Okay, I finally get it. You think that the study of human mind and behavior is split into two separate disciplines: a "medical" branch called "psychology" (not psychiatry?) that concerns itself with pathologies and a "personality" branch for things that aren't bad and nasty. Thus making value judgements a compulsory and fundamental part of studying human behavior. No wonder you're upset.
Actually, the study of personality is usually seen as a branch of psychology, the medical branch of psychology is called psychiatry, and whether to classify Asperger's Syndrome or homosexuality as pathologies, as they were regarded by the first people to refer to them as such, is up to the individual and is not inherent in the words. Psychologists are not professionally bound to disapprove of everything they study. Even psychiatrists disagree about what does and does not constitute a pathology, and in practice, it's common to allow the patient to decide which behaviors need to be modified and which behaviors the patient is happy with.
WHY, please tell me, WHY do we need a PILL for SHYNESS???
WHY do we need to talk about personalities in this medical terminologistic way???
Now I also understand why you think these two things are connected. Indeed there is a misconception, often exploited by unscrupulous doctors and businessmen, that every medical term (or psychological term, by your language) refers to a pathology that must be treated. There's even a whole set of jokes about it (most of which are probably funnier than this one):
Doctor: Mrs. Wilson, you have something called an "aorta" that originates at your heart and branches out to several other parts of your body.
Mrs. Wilson: Oh my gosh! Is it serious?
But that's nothing more than a misconception, and a fading one at that. When my dentist refers to my "upper bicuspids" I don't conclude that he's out to cure me of them. When a bodybuilder uses the word "hypertrophy," would you assume he's opposed to having bigger muscles?
It's a comical and hopeless point of view that elevates one way of seeing the world to such a position that its mere ability to speak grants it complete authority over a topic. You seem to believe that the legitimacy of the medical or psychological perspective *must* be denied, because otherwise right and wrong becomes the domain of medicine or psychology. This is exactly the kind of thinking that leaves people at the mercy of doctors who create demand for treatment by defining more and more pathologies.
Talking about something in medical terminology doesn't implicitly deprecate it. "Medical" does not mean "pathological" or even "abnormal," as you seem to think it does.
You are worried about a future where difference is pathologized and everyone is medicated into sameness. You want "MORE tolerance and acceptance of a range of quirks."
Yet you object to an article that presents Asperger's Syndrome as a normal, liveable condition that involves advantages and disadvantages and can lead to unique contributions to society.
How is this supposed to make sense? My interpretation that you were objecting to the arbitrary discretization of a spectrum was the only sensible reading I could see, so I went with it.
You have a point, but it's just the way doctors think and communicate. They don't do math; they're not scientists who deal with fields or probability distributions. They're trained to look at a patient, apply a set of labels, search through their medical school training for rules involving those labels, and then apply those rules. They investigate continuous phenomena by dividing the spectrum into discrete chunks, each with a label, and systematically creating rules involving those labels. That's why medical diagnostic systems are such a successful application of classical AI: once the doctors handle the subjective task of applying labels, the only difference between a doctor and Prolog is that Prolog can operate faster, more reliably, and using more rules than any single doctor.
Some researchers may come up with theories about continuous phenomena, but it's alien to the way doctors think.
We aren't going to lose the next generation of kids just because they're into a bunch of stupid kid things like shaking IM windows, sending Flash animations, and having cool avatars. Honestly, do you think they're still going to be doing those things when they grow up?
Linux IM clients should integrate the features that their current users find useful and not worry that today's teenagers will be just as frivolous, shallow, and glitz-oriented when they're 25 as they are now at 15. Adults have been fretting about that prospect for thousands of years, and it hasn't happened yet.
Young people have a strong bias against old and busted-looking stuff; they'll almost always favor the newest, flashiest, slickest thing. To them, Linux instant messaging must seem simply broken.
The good news is that they'll grow up and get tired of doing all the stupid stuff that only MSN enables them to do. Shaking someone's IM window is obnoxious even by teenager standards. Exchanging files is the one thing I saw in the article that I would like to do using instant messaging. Fix that (if it doesn't work already) and you'll have everyone over 18.
You can't call it "reading" and "writing." Too many computer people think that "reading" and "writing" are fast, straightforward, linear techniques that result in a disorganized mess - which for them is probably the truth. They would never understand that putting serious mental effort into creating a coherent, useful work could be called "writing," or that a systematic, intelligent effort to understand something could be called "reading."
Parent is modded Funny; I would say Insightful. I'd love to have recommender programs that know my idiosyncrasies and nudge me out of ruts and towards things that might disturb my mindset.
Hey, I do! They're called friends!
...is what's on the line here, not the security of Firefox. Installing someone else's code is always insecure if they can't be trusted. I still trust Mozilla, but I'm really shocked by this.
I agree that XML relieves the burden of parsing and unparsing, and is thus superior for any configurating file that isn't just a flat list of records. In theory, that is. The available tools just don't allow XML to live up to its potential.
I'm never tempted to use XML formats when I know I'll have to edit them. Do I edit XML in a text editor, write a custom data editor backed by XML, or write a custom parser and unparser? Usually I choose the last option. I almost never use XML for files that humans see.
Where are the editing and composing tools? Where are the XML editors that aren't slow, clunky, or oriented to a few specific document formats?
The article asserts a variety of ludicrous ideas as common conceptions about OSS. It's impossible to take seriously.
I'll grant that the point about conceptual integrity may have merit. Distributed development makes conceptual integrity very hard to maintain. But how do I know that? Through commercial experience. It only applies to OSS because almost all OSS projects are distributed.
Frankly, the ideas attributed in this article to OSS people are so alien and fantastic that I doubt the author has even read any of the basic writings about open source or studied a single open source project.
Ah, I see! It's like in calculus where you solve for f'(x) = 0 to find the minima and maxima of f. This is a perfect case study of media spin, then. Should be fun.
I think Bruce's point #5 also refers to the usefulness of code-signing in protecting against inside jobs. Since a cryptographic signature can't be faked, it can be used to hold the signer responsible for the signed content. However, as Bruce points out, signed code can let the signer off the hook by erasing its own signature. In this respect signatures on code are less useful than signatures on other kinds of content.
Software development is business, too. What do you want to bet that the WinHoldEm developers run an unreleased poker bot that has a special knack for beating WinHoldEm? It's like selling milkers to cows, then showing up with a bucket.
You don't live in a rural area. You live in the suburbs. Would the good schools and DSL would be there if it weren't for people commuting into Kansas City? Get a clue yourself.
You're right. A GENI project goal, via the article:
...) will first agree 100% that governments need absolute control and knowledge, and then they'll fight over what that control is used for.
privacy and accountability and vary protections for individuals based on "difference and local values"
"Difference and local values" is code for enabling the Saudi Government to block DanicaRacing.com.
It's also code for enabling Bush to track and harass people who read Islamic web sites, and enabling Clinton (the current one) to make sure that "parents" have absolute power to mould their "children." Bush and Clinton (and Hatch and Kennedy and
Books like that don't have to be bought and replaced often. Who cares if a compiler theory textbook is five years old, as long as the code examples aren't in Java 1.1? Leave the short-lived application-specific and language-specific stuff (except slow-changing languages like Common Lisp and historical languages like Algol) to bookstores.
Also, you should provide a way to do simple, nondangerous programming on library computers -- turtle graphics type stuff for the kiddies. You gotta make sure those future donors have fond memories :-)
Finally, you should not be disappointed that so many people here are talking about books -- you should swallow your shiny ambitions and take it to heart. It makes me sick that libraries are now pretty cool places to hang out, but I have to BRING MY OWN FRICKING BOOKS BECAUSE THE LIBRARY DOESN'T HAVE ANY. Ask yourself if anyone comes to your library for something they couldn't get at Starbucks. Your library, and libraries everywhere, may ultimately suffer the same fate as every other Starbucks competitor.
This is why I skip to the comments when an article sounds too good to be true. Thanks for saving me five minutes.
Even as someone who abandoned Perl for Python, I suspect there is a point to Perl. I'm truly impressed by the concise power of Perl one-liners and text-manipulation scripts, and I made an earnest attempt to become a Perl user.
:-)
Unfortunately, I only need scripting once every few months or weeks, which is longer than it takes me to forget all the little details that make Perl so rich and expressive. Python is much better for me: it only takes me a few minutes to start writing good code in Python, no matter how long it's been since the last time.
My failure should give some idea of who can and should use Perl:
1) People who need and use scripting every day or at least several days per week
2) People who are much smarter than me
Seriously, Perl is a complex and sophisticated tool that rewards consistent commitment. You just can't half-ass it, or neglect your Perl skills and expect them to be there months later when you need them. For that, use Python.