I've often considered the possibility that technical/scientific people tend to apply overly simplistic models to complex real-life systems.
I find myself trying to learn social topics by forming crude rule-systems around them. This seems to drive my wife crazy until I happen upon a set of rules that matches her (experienced or heard-of) evidence and anecdotes. She's the feedback mechanism in the system.
Ok, I'll bite. Concrete example: I've seen BOOST_FOREACH cause my compiler to fail to optimize a std::vector iteration down to the appropriate pointer arithmetic. I replaced it with a simple for loop and the problem went away. If memory serves, this was GCC 4.3.x and boost 1.40. Theoretical argument: In my experience (blogs and conversations), proponents of the "modern C++" approach (boost or any other heavy dependency on templates/inlining) tend to hand-wave the compiler's optimization pass. In a large codebase, small changes in inline functions can have chaotic effects on an optimizer's heuristics. Because the code is inline, there is no way for a library developer to reliably tune for good performance for a downstream developer. Furthermore, unnecessary duplication in template logic (mostly from poorly-optimized template classes) can bloat the instruction stream something awful, which will kill cache performance.
Also, boost adds a lot of compilation overhead, and I don't have a sword at my desk. Most libraries at least incorporate config.hpp, which ends up bringing in <cmath> and messes with any code that was already using the (global namespace) macros defined in <math.h>. IIRC same with several other standard headers: the "C++" version is not entirely compatible with users of the "C" version, even on the most popular platform (glibc/linux).
Make it illegal for regulators to ever enter/return to the industry they regulate. Hire experts in the industry and/or train people to become experts in the industry and pay them fair wages, rather than relying on a system where there's a revolving door and money rains down on those who come back to the private sector.
I think it makes sense to hire regulators from the industry, because they will know and really understand how it works from the competitive side. You can't teach that on the regulatory side. However, something similar to a several-year non-compete (and enforceable in all states) is definitely appropriate to reduce the incentive for bribery and regulatory capture.
Otherwise, campaign for a third party (promote it, whatever), but vote for the lesser of the two big evils.
That certainly helps, but votes carry much more teeth than campaigns. Some small percentage of the population has to be willing to "throw away" their vote to an independent (or Green, Libertarian, Socialist, etc.) in order to convince the larger population to start taking the party seriously and stop voting for what they perceive as the second-worst candidate.
Be the early adopter that you want to see in the world.
[Copyright is] not any more or less artificial than many common modes of private property, such as stocks.
Are you suggesting that stocks are artificial? They represent beneficial ownership which is very real. Furthermore, the paper or electronic records that they reside in are no less real than fiat currency. Copyright (or any other "IP" right) is an institutionalized monopoly, not like property at all.
Oh, and higher taxes (33% from income over US$63k).
To be honest, that doesn't sound that high. I don't know what 63K USD gets in terms of lifestyle, so 33% may be quite a bit for the middle class. But it's paltry for high-income earners, just the same as the US tax system. Also, income tax is just a small piece of the puzzle: US long-term capital gains (e.g. many investments held at least 1 year) cap out much lower (IIRC 15%) and represent more of the "income" of wealthy individuals/organizations. Plus, capital losses can be applied to later gains to offset tax burden. This is convenient in contrast to Joe Schmoe whose "gains" (income) and "losses" (i.e. income-interrupting events) both tend to be taxable (disability, workman's comp, unemployment) or breakeven (unpaid time off, maybe furlough).
So how do long-term capital gains get taxed in NZ?
Your ass propels you with as much force as it propels others.
But that's only force, and we all know F==m*a. So his ass must propel his mass, while the gas from his ass must propel the mass of the masses behind his ass.
"You're awkward and terrible with women, therefore you're a sexist."
I agree with you that (glorified) awkwardness does not imply sexism. However, I don't think the GP was far off the mark: I knew guys in college that would regularly and unintentionally creep out girls in casual conversation. They just didn't catch onto their own body language. I can see that alienating any girls that are not equally awkward. But it's not sexism, not by a long shot.
// Somebody pasted a few lines, but didn't reindent last line. // Still correct syntax, still does the same, Reindent All fixes the formatting and everybody's happy. if (x) { statement1 statement2 statement3 }
If I could write a pre-commit hook that rejected this based on inconsistent indentation, I would. Code with consistently inconsistent formatting drives me batty!
I'm a little out of my depth here, but I'd imagine it'd be much easier with UDP because UDP is connectionless. With this sort of roaming, the server isn't expected to change addresses, but the client is. So, have the client sign everything with a negotiated public key, and the server doesn't even have to care where each packet is coming from, or even open any new connections when the client moves across IPs.
In the context of persistent logical connections, you have to consider that the TCP connection will get severed and must be re-established. It's not enough to have an open connection sitting around; if bits aren't coming from the other side, the open connection may have been physically severed by some dead networking gear or a backhoe. So you end up coding some application-level heartbeat-and-reconnect logic with logic to securely resume a session (forgive the hand-waving here).
At this point, if the "securely resume a session" bit is sufficiently compact, you are free to do the exact same logic per-packet, and you don't need to maintain a TCP connection. You lose TCP's reliable-ordering support, but the automatic-reconnect logic is typically sufficient to do without reliable-ordering. You also lose TCP's congestion control, so you'll probably need to add some application-level throttling mechanism to avoid DoSing yourself.
This is why TCP vs. UDP is irrelevant in the topic at hand. Neither is sufficient to maintain a persistent connection, and the extra logic required on top of one can typically be ported to the other.
Buying stock is a risk reward proposition. I own a portion of the company equal to the number of shares I own over the total number of shares available.
If you're buying common stock (which is most stocks), you're also the last in line to get paid when the company declares bankruptcy. You can still lose everything with no recourse, it's just less common because the barrier to entry is higher for publicly-owned companies than Kickstarter projects.
In short, Ron Paul is really just another side of the same coin as the Democrats and the Republicans. That's not what we need. What we need is to throw away the rusty old coin entirely and bring in people with fresh ideas.
Given the current theater of candidates, Paul is the fresh idea. He's not young, but he's different enough that a big voting block would scare other candidates and local lawmakers into attentiveness (to the constituency rather than lobbying interests).
you're entitled to that opinion, and nothing more. that's the issue here. bioware has no obligation to change what you don't like.
True, but that's not the full story. No, you will probably not get very far by demanding a refund from Bioware or by threatening to force them (via lawsuit) to change the ending. But most people complaining are people who have purchased Bioware games in the past, and are likely to purchase them in the future. The company is absolutely free to alienate its user base; it's their funeral. The free DLC is an attempt to make amends precisely because many people voiced their opinions in forums like this.
...and even when they're not, just prepend !g to the query, resubmit, and you get Google's search results via encrypted.google.com with no obnoxious auto-complete for partial search queries. Nowadays I only visit the Google homepage on exciting holidays like Les Paul's birthday.
I remember thinking, "Well, that's bizarre". It completely ruined the song.
I have an Audyssey-enabled receiver, which does similar normalization. I remember this same effect on the intro to Coheed & Cambria's "Welcome Home". Normalization totally destroys the impact. OTOH, it's good for providing a more consistent noise floor when company is over.
Up until late last year [ALAC] was proprietary, but as of last October it is now Open Source under the Apache 2.0 license. The compression level is on par with FLAC.
IIRC ALAC came after FLAC (Wikipedia confirms this: FLAC 2001, ALAC 2004)... but Apple went and created their own proprietary format that is par and not better? Seems pretty antisocial unless there is something about ALAC that makes it substantially more efficient to decode. Open-sourcing it after 7 years just sounds like more empire-building to me.
In my experience with on-stage monitors, each musician has only wanted to hear themselves louder.
That doesn't sound surprising. Hearing yourself perform provides vital feedback to a musician. In an entirely-amplified environment, it's too easy to just lose yourself in a mix, especially in a supporting (i.e. quieter to the audience) role.
I've got this song on my iPhone called "Map of the Problematique" by Muse, it has almost no dynamic range... That song is a masterpiece.
I have that song. It's awesome in my car. It sounds awful in my basement; the constant "crushed" feeling distracts too much from the overall energy of the song. My present solution is to listen to Foo Fighters at home and Muse on the road, but I really wish that someone would remaster BH&R. It is engineered worse than both their earlier and later albums, but IMHO it contains the best material, so I think it would sell.
It's sad that music lovers have to run to a video game to get the best source material though.
The same thing happened for Muse's Knights of Cydonia: the Guitar Hero track was used for a community remaster. The entirety of Black Holes and Revelations sounds quite nasty through a real sound system, which is sad given the awesomeness of the material.
Which usually is the type of thing you simply cannot do in Python, such as deliberate aliasing to avoid copies or known downcasts based on meatspace knowledge (i.e. compiler can't optimize away the RTTI lookup for dynamic_cast, but programmer knows that static_cast will always succeed because of some external constraint). But that is neither here nor there. Your assertion appears to be that Python has safe dynamic run-time type support, and C++ only has raw pointer casts. Yet, C++ specifically added the features I mentioned so that programmers could get more of the "Python" safety you espouse, and less finger-crossing with C-style casts. But the old behavior still exists for backwards compatibility and for extreme optimization (because really, dynamic_cast shouldn't ever be the bottleneck). So that is why I am calling your argument a strawman: you are ignoring critical language features in order to make your point! After all, one could also cherry-pick the worst of Python: code which doesn't handle duck-typing gracefully, in particular anything that assumes values are floating-point numbers without doing anything to "coerce" them. The whole thing with/ vs.// in Python3 is focused on this particular issue.
Your argument would be better suited against C. C++ is a different beast.
Slashdot Mirror
I find myself trying to learn social topics by forming crude rule-systems around them. This seems to drive my wife crazy until I happen upon a set of rules that matches her (experienced or heard-of) evidence and anecdotes. She's the feedback mechanism in the system.
Ok, I'll bite. Concrete example: I've seen BOOST_FOREACH cause my compiler to fail to optimize a std::vector iteration down to the appropriate pointer arithmetic. I replaced it with a simple for loop and the problem went away. If memory serves, this was GCC 4.3.x and boost 1.40. Theoretical argument: In my experience (blogs and conversations), proponents of the "modern C++" approach (boost or any other heavy dependency on templates/inlining) tend to hand-wave the compiler's optimization pass. In a large codebase, small changes in inline functions can have chaotic effects on an optimizer's heuristics. Because the code is inline, there is no way for a library developer to reliably tune for good performance for a downstream developer. Furthermore, unnecessary duplication in template logic (mostly from poorly-optimized template classes) can bloat the instruction stream something awful, which will kill cache performance.
Also, boost adds a lot of compilation overhead, and I don't have a sword at my desk. Most libraries at least incorporate config.hpp, which ends up bringing in <cmath> and messes with any code that was already using the (global namespace) macros defined in <math.h>. IIRC same with several other standard headers: the "C++" version is not entirely compatible with users of the "C" version, even on the most popular platform (glibc/linux).
I think it makes sense to hire regulators from the industry, because they will know and really understand how it works from the competitive side. You can't teach that on the regulatory side. However, something similar to a several-year non-compete (and enforceable in all states) is definitely appropriate to reduce the incentive for bribery and regulatory capture.
Genetic supremacy? And you're posting on /.?
That certainly helps, but votes carry much more teeth than campaigns. Some small percentage of the population has to be willing to "throw away" their vote to an independent (or Green, Libertarian, Socialist, etc.) in order to convince the larger population to start taking the party seriously and stop voting for what they perceive as the second-worst candidate.
Be the early adopter that you want to see in the world.
Are you suggesting that stocks are artificial? They represent beneficial ownership which is very real. Furthermore, the paper or electronic records that they reside in are no less real than fiat currency. Copyright (or any other "IP" right) is an institutionalized monopoly, not like property at all.
To be honest, that doesn't sound that high. I don't know what 63K USD gets in terms of lifestyle, so 33% may be quite a bit for the middle class. But it's paltry for high-income earners, just the same as the US tax system. Also, income tax is just a small piece of the puzzle: US long-term capital gains (e.g. many investments held at least 1 year) cap out much lower (IIRC 15%) and represent more of the "income" of wealthy individuals/organizations. Plus, capital losses can be applied to later gains to offset tax burden. This is convenient in contrast to Joe Schmoe whose "gains" (income) and "losses" (i.e. income-interrupting events) both tend to be taxable (disability, workman's comp, unemployment) or breakeven (unpaid time off, maybe furlough).
So how do long-term capital gains get taxed in NZ?
FTFY
But that's only force, and we all know F==m*a. So his ass must propel his mass, while the gas from his ass must propel the mass of the masses behind his ass.
I agree with you that (glorified) awkwardness does not imply sexism. However, I don't think the GP was far off the mark: I knew guys in college that would regularly and unintentionally creep out girls in casual conversation. They just didn't catch onto their own body language. I can see that alienating any girls that are not equally awkward. But it's not sexism, not by a long shot.
If I could write a pre-commit hook that rejected this based on inconsistent indentation, I would. Code with consistently inconsistent formatting drives me batty!
In the context of persistent logical connections, you have to consider that the TCP connection will get severed and must be re-established. It's not enough to have an open connection sitting around; if bits aren't coming from the other side, the open connection may have been physically severed by some dead networking gear or a backhoe. So you end up coding some application-level heartbeat-and-reconnect logic with logic to securely resume a session (forgive the hand-waving here).
At this point, if the "securely resume a session" bit is sufficiently compact, you are free to do the exact same logic per-packet, and you don't need to maintain a TCP connection. You lose TCP's reliable-ordering support, but the automatic-reconnect logic is typically sufficient to do without reliable-ordering. You also lose TCP's congestion control, so you'll probably need to add some application-level throttling mechanism to avoid DoSing yourself.
This is why TCP vs. UDP is irrelevant in the topic at hand. Neither is sufficient to maintain a persistent connection, and the extra logic required on top of one can typically be ported to the other.
If you're buying common stock (which is most stocks), you're also the last in line to get paid when the company declares bankruptcy. You can still lose everything with no recourse, it's just less common because the barrier to entry is higher for publicly-owned companies than Kickstarter projects.
Given the current theater of candidates, Paul is the fresh idea. He's not young, but he's different enough that a big voting block would scare other candidates and local lawmakers into attentiveness (to the constituency rather than lobbying interests).
True, but that's not the full story. No, you will probably not get very far by demanding a refund from Bioware or by threatening to force them (via lawsuit) to change the ending. But most people complaining are people who have purchased Bioware games in the past, and are likely to purchase them in the future. The company is absolutely free to alienate its user base; it's their funeral. The free DLC is an attempt to make amends precisely because many people voiced their opinions in forums like this.
If you're using a multicore server, you can use CPU affinity (taskset in many linux distros) to keep your TF2 and minecraft servers segregated.
Even longer, if you look at castle wall construction.
I have an Audyssey-enabled receiver, which does similar normalization. I remember this same effect on the intro to Coheed & Cambria's "Welcome Home". Normalization totally destroys the impact. OTOH, it's good for providing a more consistent noise floor when company is over.
IIRC ALAC came after FLAC (Wikipedia confirms this: FLAC 2001, ALAC 2004)... but Apple went and created their own proprietary format that is par and not better? Seems pretty antisocial unless there is something about ALAC that makes it substantially more efficient to decode. Open-sourcing it after 7 years just sounds like more empire-building to me.
That doesn't sound surprising. Hearing yourself perform provides vital feedback to a musician. In an entirely-amplified environment, it's too easy to just lose yourself in a mix, especially in a supporting (i.e. quieter to the audience) role.
I have that song. It's awesome in my car. It sounds awful in my basement; the constant "crushed" feeling distracts too much from the overall energy of the song. My present solution is to listen to Foo Fighters at home and Muse on the road, but I really wish that someone would remaster BH&R. It is engineered worse than both their earlier and later albums, but IMHO it contains the best material, so I think it would sell.
The same thing happened for Muse's Knights of Cydonia: the Guitar Hero track was used for a community remaster. The entirety of Black Holes and Revelations sounds quite nasty through a real sound system, which is sad given the awesomeness of the material.
Which usually is the type of thing you simply cannot do in Python, such as deliberate aliasing to avoid copies or known downcasts based on meatspace knowledge (i.e. compiler can't optimize away the RTTI lookup for dynamic_cast, but programmer knows that static_cast will always succeed because of some external constraint). But that is neither here nor there. Your assertion appears to be that Python has safe dynamic run-time type support, and C++ only has raw pointer casts. Yet, C++ specifically added the features I mentioned so that programmers could get more of the "Python" safety you espouse, and less finger-crossing with C-style casts. But the old behavior still exists for backwards compatibility and for extreme optimization (because really, dynamic_cast shouldn't ever be the bottleneck). So that is why I am calling your argument a strawman: you are ignoring critical language features in order to make your point! After all, one could also cherry-pick the worst of Python: code which doesn't handle duck-typing gracefully, in particular anything that assumes values are floating-point numbers without doing anything to "coerce" them. The whole thing with / vs. // in Python3 is focused on this particular issue.
Your argument would be better suited against C. C++ is a different beast.
This is a strawman. This is a language feature for portability with C that only shows up in poorly-written C++ programs.