Mathematica has more C code than you realize. According to wolfram the kernel is 1.5 million lines of C and 150k lines of mathematica.
The growth of the C code base has been necessary to compete with products like matlab that have lots of fast machine-precision algorithms.
According to wolfram "a fraction of a percent" of the C is architecture-specific. That could easily be some thousands of lines. The only reason they are able to port to x86 in 2 hours is that they have already done it.
FWIW the powermac has always done poorly on the mathematica benchmarks... a top-of-the-line dual g5 is about half the speed of a top amd or intel system. So I think its a good thing they are switching because I like mathematica on OSX but its a currently a poor choice if you need a high power system.
studies have found that people who did a lot of binge drinking when they were kids develop a kind of resistance to alchohol whereby they are immune to its depressive effects (i.e. they retain their adolescent response). they are also at least an order of magnitude more likely to be chronic alcoholics as adults.
this change in brain chemistry explains why many severe alchoholics can drink all day and still function normally, it also validates the grandparents assertion to some degree--its quite possible that his brain chemistry has a greater resistence to alchohol (not that this excuses one from driving over the limit but who knows, it might be admissible evidence in a trial given an objective test).
simple solution; put a bin in the smoking area where you can drop off your lighter prior to entry. when you come out on the other side, there will be another bin where you can pick up someone else's discarded lighter.
this is an editorial piece, nothing scientific about it. there are no references to the supposed studies they quote, there are not even second opinions from authors or other scientific researchers. pure speculation.
and many those research problems cited are *not* obvious, but the author belittles the studies they quote to make the research questions sound more obvious than they really are.
* on a finding that men over 55 are a high risk group for digit loss due to power tools... this requires good statistics to know, its *not* obvious.
* that workers are less efficient in a cold environment... again, not obvious, and many workplaces (imho) keep the thermostat too low.
* that asthma and smoking aggregate worse than either, again not obvious... many people falsely believe this is not true.
* that doctor-patient communication is critical for reducing harmful effects of mixing drugs, also not obvious -- now we know that communication skills are an important part of medical training.
granted, many of the studies conclude with obvious recommendations, e.g., "be careful with power tools", and the author makes great fun of these "obvious conclusions" when in fact, they are not the substantiative conclusion (i.e., factual finding) of the research, just a recommendation for how to interpret the finding.
another method if you don't have jets (which is a cool idea, btw); submerge the hose to fill it with water (might take a bit of finesse since there is a lot of air trappen inside), then cap one end with the palm of you hand and pull it out.
*sigh* trade wars... i used to be obsessed with that game. at first, it started by installing scripts to automate trading... then I started exploiting the plethora of bugs to become super powerful, especially the "clone planet" bug (which also cloned all the money in the planet's bank), then I would buy a zillion probes and fire them off into every sector of the galaxy (again, scripted to do this) and blow away every bad guy until I could buy the Enterprise, at which point I would turn to the dark side (the bad guys can't buy the enterprise), do the probe sweep again and blow away every other player in the galaxy.. I guess you could say I was a total brat.
spam zombies are only the first sign of trouble. next comes ssh zombies, DoS zombies, keystroke loggers/phishers, ransom-for-data lockouts, etc. not only are they a general nusiance but the owners need some sort of painfully obvious notification that they have been 0wned.
mathematica's numeric capability has vastly increased with version 5, which makes it a strong matlab competitor. mathematica expressions can also be compiled to remove the type polymorphism and interpreter overhead.
the upside of numpy is that its python, so you have the whole python foundation to work with... custom GUIs, database connections, web services, whatever you want, its there, its free, and its relatively easy to write new C modules if you need them.
however things get kind of scattered when you try to do a big project. you will find out that what you need is module X, which is part of some bigger project, e.g. scipy, python-scientific, pysparse, etc etc (there are several). as you start to combine more and more modules from different places, the code starts to get messy because you are mixing all these apis together.
and yes, plotting is pretty bad -- py-gnuplot is the best available, and it will only get you so far. some people try to integrate with blender or some GL toolkit to get better graphics... YMMV.
I used to think the open-source ethic of using numpy outweighed the usefulness and completeness of commercial tools, but I don't believe that any more (though it may still be true in some cases). if you are in academics you can get matlab at home for cheap, and if you are an engineer and need it for work, then the high price is not really an issue.
if you are doing research its probably in an academic environment, and probably a competetive one, and if you use numpy the matlab folks are going to code circles around you.
all of that said I prefer to use mathematica -- which is by now fairly robust for fast numeric work as to be competetive with matlab, but with a lot more capabilities -- but its a *significantly* more difficult language/api to master... heavy emphasis on functional programming, lots of non-standard operators, etc.
even if we suppose that the complexity of DNA exceeds that which can be described in evolutionary theory, that does not give any endorsement for "god" (no matter how minimal). "super-intelligent wisdom begat energy" is still a non-sensical, non-testable assertion. if it was indeed provable that DNA exceeds evolution (and believe me, my world would not be shattered if this was found to be the case), it would be rather easy to come up with a number of testable scientific hypothesis to explain it -- there is no need to resort to Schroeder's nonsense.
...and the giant snail with a transparant shell that transported him across the ocean? That was one cool snail. Even though my ADSL has 6mbit downstream, I'd trade a snail ride for my line anyday.
it kind of depends on where you are, though. yellow lights in san francisco city are incredibly long. if you just stopped as soon as you saw yellow, you'd have some pissed off people behind you.
but out in rural areas the yellow tend to be pretty short.
its a valid nitpick, but with all the photo enforced lights getting installed lately, I think people may finally start to respect the signals.
actually thats right that darpa thinks the course is reasonably navigable; but there was more than one person who thought the judges were... misjudging.
darpa grand challenge is kind of an extreme case for a few reasons... not only is it a fairly tough AI problem, there are some incredible engineering challenges as well.
- its not clear that even a human driver would be able to do the course in the given amount of time. this is *difficult* off road driving, even for a professional off-road racer. the probability of getting stuck or breaking an axel or other mechanical failure is non-trivial.
- driving off road makes the sensor reading problem about 100 times worse due to the extreme vibration which is translated to the sensors. last year the cmu team was the favorite, in no small part due to the incredibly expensive shock isolation mount on top of the vehicle (one of which was destroyed in the early trials when the hummer flipped over)
doing the same thing for highway driving in normal traffic is relatively easy, in fact there are already prototypes that can do it.
(according to my relatively uninformed theory) basically you can think of the the visual processing system as a set of imaging filters connected to a collection of non-linear discriminators. the filters are intrinsically geometric because that is the optimal configuration for multi-aspect target recognition -- an optimal pattern reached by evolution, learning or some combination thereof. (its also interesting to note that the facial recognition subsystem is -not- multi-aspect, it only works when faces are oriented upwards).
thus, drugs don't directly cause geometric hallucinations, they simply cause the system to malfunction slightly in such a way that the underlying structure is revealed (kind of like feeding an impulse into a filter to estimate its frequency response -- the malfunctions consist of a bunch of random activity spikes were none are normally expected).
this is not to contradict what you are saying, which I think is a neat idea. but it may suggest that the development is not unique (i.e., the optimality of geometric filters is universal or mathematical in nature)
personally i feel noticably stupider for 1-2 days after smoking pot, and then its gone. hallucinogens make me feel like a genius for about 48 hours, and then when it wears off I can't remember what was so great about my ideas.
its an app layer problem all right, and one that has become an epidemic amoung the hoards of naive php scripters, but a simple one to fix. db drivers should provide their own string substitution methods that apply proper quoting... (e.g., db::sprintf)... the python db api does an excellent job of providing this.
Slashdot Mirror
Mathematica has more C code than you realize. According to wolfram the kernel is 1.5 million lines of C and 150k lines of mathematica.
The growth of the C code base has been necessary to compete with products like matlab that have lots of fast machine-precision algorithms.
According to wolfram "a fraction of a percent" of the C is architecture-specific. That could easily be some thousands of lines. The only reason they are able to port to x86 in 2 hours is that they have already done it.
FWIW the powermac has always done poorly on the mathematica benchmarks... a top-of-the-line dual g5 is about half the speed of a top amd or intel system. So I think its a good thing they are switching because I like mathematica on OSX but its a currently a poor choice if you need a high power system.
According to the Mathematica 5 manual, there "several million" tests that take at least "a few days" to run.
Verification of the port in 2 hours would likely require a 64-node cluster.
interesting.
fwiw my crt has convergence controls in the menu (its an iiyama 22")
studies have found that people who did a lot of binge drinking when they were kids develop a kind of resistance to alchohol whereby they are immune to its depressive effects (i.e. they retain their adolescent response). they are also at least an order of magnitude more likely to be chronic alcoholics as adults.
this change in brain chemistry explains why many severe alchoholics can drink all day and still function normally, it also validates the grandparents assertion to some degree--its quite possible that his brain chemistry has a greater resistence to alchohol (not that this excuses one from driving over the limit but who knows, it might be admissible evidence in a trial given an objective test).
simple solution; put a bin in the smoking area where you can drop off your lighter prior to entry. when you come out on the other side, there will be another bin where you can pick up someone else's discarded lighter.
rtfa;
this is an editorial piece, nothing scientific about it. there are no references to the supposed studies they quote, there are not even second opinions from authors or other scientific researchers. pure speculation.
and many those research problems cited are *not* obvious, but the author belittles the studies they quote to make the research questions sound more obvious than they really are.
* on a finding that men over 55 are a high risk group for digit loss due to power tools... this requires good statistics to know, its *not* obvious.
* that workers are less efficient in a cold environment... again, not obvious, and many workplaces (imho) keep the thermostat too low.
* that asthma and smoking aggregate worse than either, again not obvious... many people falsely believe this is not true.
* that doctor-patient communication is critical for reducing harmful effects of mixing drugs, also not obvious -- now we know that communication skills are an important part of medical training.
granted, many of the studies conclude with obvious recommendations, e.g., "be careful with power tools", and the author makes great fun of these "obvious conclusions" when in fact, they are not the substantiative conclusion (i.e., factual finding) of the research, just a recommendation for how to interpret the finding.
another method if you don't have jets (which is a cool idea, btw); submerge the hose to fill it with water (might take a bit of finesse since there is a lot of air trappen inside), then cap one end with the palm of you hand and pull it out.
*sigh* trade wars... i used to be obsessed with that game. at first, it started by installing scripts to automate trading... then I started exploiting the plethora of bugs to become super powerful, especially the "clone planet" bug (which also cloned all the money in the planet's bank), then I would buy a zillion probes and fire them off into every sector of the galaxy (again, scripted to do this) and blow away every bad guy until I could buy the Enterprise, at which point I would turn to the dark side (the bad guys can't buy the enterprise), do the probe sweep again and blow away every other player in the galaxy.. I guess you could say I was a total brat.
spam zombies are only the first sign of trouble. next comes ssh zombies, DoS zombies, keystroke loggers/phishers, ransom-for-data lockouts, etc. not only are they a general nusiance but the owners need some sort of painfully obvious notification that they have been 0wned.
hm, interesting.... but do you think they struggle because of language limitation or because they are inexperienced programmers?
mathematica's numeric capability has vastly increased with version 5, which makes it a strong matlab competitor. mathematica expressions can also be compiled to remove the type polymorphism and interpreter overhead.
i've done a fair amount of work in numpy --
the upside of numpy is that its python, so you have the whole python foundation to work with... custom GUIs, database connections, web services, whatever you want, its there, its free, and its relatively easy to write new C modules if you need them.
however things get kind of scattered when you try to do a big project. you will find out that what you need is module X, which is part of some bigger project, e.g. scipy, python-scientific, pysparse, etc etc (there are several). as you start to combine more and more modules from different places, the code starts to get messy because you are mixing all these apis together.
and yes, plotting is pretty bad -- py-gnuplot is the best available, and it will only get you so far. some people try to integrate with blender or some GL toolkit to get better graphics... YMMV.
I used to think the open-source ethic of using numpy outweighed the usefulness and completeness of commercial tools, but I don't believe that any more (though it may still be true in some cases). if you are in academics you can get matlab at home for cheap, and if you are an engineer and need it for work, then the high price is not really an issue.
if you are doing research its probably in an academic environment, and probably a competetive one, and if you use numpy the matlab folks are going to code circles around you.
all of that said I prefer to use mathematica -- which is by now fairly robust for fast numeric work as to be competetive with matlab, but with a lot more capabilities -- but its a *significantly* more difficult language/api to master... heavy emphasis on functional programming, lots of non-standard operators, etc.
Asbestos is safe? Care to back that statement up with some facts?
definately X if you include those abu ghraib scenes...
even if we suppose that the complexity of DNA exceeds that which can be described in evolutionary theory, that does not give any endorsement for "god" (no matter how minimal). "super-intelligent wisdom begat energy" is still a non-sensical, non-testable assertion. if it was indeed provable that DNA exceeds evolution (and believe me, my world would not be shattered if this was found to be the case), it would be rather easy to come up with a number of testable scientific hypothesis to explain it -- there is no need to resort to Schroeder's nonsense.
...and the giant snail with a transparant shell that transported him across the ocean? That was one cool snail. Even though my ADSL has 6mbit downstream, I'd trade a snail ride for my line anyday.
Slashdot comments are more important than....... uh, I can't think of anything right now. Anyone?
it kind of depends on where you are, though. yellow lights in san francisco city are incredibly long. if you just stopped as soon as you saw yellow, you'd have some pissed off people behind you.
but out in rural areas the yellow tend to be pretty short.
its a valid nitpick, but with all the photo enforced lights getting installed lately, I think people may finally start to respect the signals.
hey, beetles are pretty smart.
actually thats right that darpa thinks the course is reasonably navigable; but there was more than one person who thought the judges were... misjudging.
darpa grand challenge is kind of an extreme case for a few reasons... not only is it a fairly tough AI problem, there are some incredible engineering challenges as well.
- its not clear that even a human driver would be able to do the course in the given amount of time. this is *difficult* off road driving, even for a professional off-road racer. the probability of getting stuck or breaking an axel or other mechanical failure is non-trivial.
- driving off road makes the sensor reading problem about 100 times worse due to the extreme vibration which is translated to the sensors. last year the cmu team was the favorite, in no small part due to the incredibly expensive shock isolation mount on top of the vehicle (one of which was destroyed in the early trials when the hummer flipped over)
doing the same thing for highway driving in normal traffic is relatively easy, in fact there are already prototypes that can do it.
(according to my relatively uninformed theory) basically you can think of the the visual processing system as a set of imaging filters connected to a collection of non-linear discriminators. the filters are intrinsically geometric because that is the optimal configuration for multi-aspect target recognition -- an optimal pattern reached by evolution, learning or some combination thereof. (its also interesting to note that the facial recognition subsystem is -not- multi-aspect, it only works when faces are oriented upwards).
thus, drugs don't directly cause geometric hallucinations, they simply cause the system to malfunction slightly in such a way that the underlying structure is revealed (kind of like feeding an impulse into a filter to estimate its frequency response -- the malfunctions consist of a bunch of random activity spikes were none are normally expected).
this is not to contradict what you are saying, which I think is a neat idea. but it may suggest that the development is not unique (i.e., the optimality of geometric filters is universal or mathematical in nature)
personally i feel noticably stupider for 1-2 days after smoking pot, and then its gone. hallucinogens make me feel like a genius for about 48 hours, and then when it wears off I can't remember what was so great about my ideas.
its an app layer problem all right, and one that has become an epidemic amoung the hoards of naive php scripters, but a simple one to fix. db drivers should provide their own string substitution methods that apply proper quoting... (e.g., db::sprintf)... the python db api does an excellent job of providing this.
i think you meant to say "performance" not "availablity"...
and fwiw mysql is only high performance in read-mostly non-transactional environments.
data corruption can also be caused by lack of sufficient features.
case in point, journaled vs. non-journaled filesystems, or the RDBMS equivalent which is write-ahead-logging (e.g., postgres).
microsoft announced today their new marketing campaign titled "switch back"