Are you saying that in America you can get a criminal record for littering?
It's up to the state, but certainly you can in some states (maybe all states, I don't know). Certainly it's on your record in California and Texas, to name a couple big states at opposite ends of the political spectrum.
Not every crime is put on your record; the so-called "infractions", which are less severe than misdemeanors or felonies, are sometimes omitted. Not always, though--speeding tickets are often (lower-than-misdemeanor) infractions, but they're certainly tracked for at least a few years in most areas.
If you take it out when it hits 155 F, the internal temperature can then get up to 165 before it starts cooling down. Now granted, it is unlikely to get much over 160, but it's probably going to taste better and should still be perfectly safe.
You get a lot more carryover heat gain than you'd expect.
Alton Brown's show this year he's revised in a couple of ways: 1. No breast-plate foil any more 2. Pull the bird out when the breast reaches 151F.
He did that with a 12 lb bird--the breast was 151F when he pulled it out, and after 1/2 hour of rest the carryover had raised the breast temp to 164F.
An in other areas of the U.S. they have moved enforcement from criminal to civil. Here in RI you can have up to an ounce, you can't smoke in public, and they actually license growers. It's all very quiet of course but it exists.
If you are caught in public it's like a $100 fine on the civil side, not criminal.
I hate the term "decriminalization" and wish it would go away, because it leads to confusion like this.
In RI and MA, the penalty for marijuana possession is now only a fine (no jail time or arrest). But it's still a criminal fine, not a civil fine. Criminal fines are punitive; civil fines are purely restitution for actual damages done to the government. If you dump toxic waste in a national park, you might be assessed a civil fine to pay for the costs of cleanup (on top of any possible criminal penalties for trespass, endangering people's health, etc). If you litter or speed, you can be assessed a criminal fine. Even though in common parlance people use "criminal" to mean misdemeanors and felonies, strictly speaking infractions like littering and speeding are still (very minor) crimes, and it's because of that misuse that the misleading term "decriminalization" came into play.
In general, a good way to differentiate "civil" cases (torts) vs. "criminal" cases (crimes) is to look at who the non-defendant is in the case: If you fail to fulfill the terms of a contract, or file for divorce, or are sued for child support, then the other party will hire a private attorney and sue you in civil court. If you assault someone, rape someone, commit DUI, etc, then the state's attorney (not some attorney hired by the victim) will show up and take you to a criminal court. There are many levels of criminal courts, ranging from traffic courts up to the US Supreme Court, alongside the civil court system.
The case of a "civil fine" is more complicated, but it's essentially the case where the government is showing up not to protect the general public or a third party, but to recover damage to government property. It's similar to a true civil case in that the wronged party is representing themselves in court, but in practice that's tough to distinguish because the wronged party is the government.
Yeah, the leading question is a little annoying--it should be "does anything keep you from using Linux, and if so what?" My answer would be the same as yours--I've been running Linux exclusively for 18 years now. For others, there are still applications--games or otherwise--or usability issues, driver support, or whatever that have stopped them from adopting Linux.
It's the same as the annoying "Will this be the year of the Linux desktop?" questions--the closest thing we'll ever see to a "year of the Linux desktop" was around 2000 or 2001, when you started seeing Dell and others selling Linux machines, could actually use a graphical "click here" widget to set up networking instead of needing to manually enter your gateway and network addresses by hand, didn't need to manually enter your monitor's timings as XF86 modelines, etc.
That's a far cry from saying that it was usable on the desktop by everyone, but it's basically the closest thing to a bright line we're ever liable to see, and is the point where it crossed over from "power users can use a Linux desktop if true gurus install it for them" to "regular users can use it if power users install it". From there on out, it's incremental support for apps here and there, UI improvements, driver support, etc--but there will probably never be another year of the Linux desktop as big as what we already saw 10+ years ago, it'll be slow adoption in fits and starts here and there as people's particular needs are met. The biggest global hurdles are long since passed, and we're into an (equally important, but harder to solve at one whack) forest of individual issues holding things back for people.
You're better off running your own CA and distributing that CA's public key to your internal apps. Then you can ignore outside CAs but still avoid MITM attacks.
Canada is party to both agreements. The US is party to the London Convention. Russ George is an American, his company is American, and they were working for the Haida (a Canadian Aboriginal group) so they are in legal trouble if the Canadian Courts find either applies
More accurately, they _might_ be in trouble if Canadian Courts find that either applies. It'll then depend on the ongoing mess of sovereignty issues with the Council of the Haida Nations--the Haida never signed any treaties with the Canadian government, so they have a much greater claim to sovereignty than many of the First Nations and may have a legal argument that they are not signatory to those treaties (despite Canada being a signatory).
The hardest thing mixed in there was the a.out->ELF migration, for which I rebuilt everything on the system by hand sort of like a primitive LFS. It was worth doing once--you learn a fair bit in the process--but made it so I have little interest in LFS or similar distros going forward.
To be fair, the NSA don't seem to have caused problems with the S-Boxes and differential analysis doesn't seem to have worked too well
In fact, the NSA's changes to the S-boxes made DES stronger against differential cryptanalysis; it appears that they and IBM knew about diff crypto back in the 1970s and designed an algorithm to resist it even though the technique wouldn't be widely known for another 15-20 years.
Diffential crypto only "doesn't seem to have worked out so well" because it's known and algorithms are designed to withstand it--prior to the public knowledge of diff crypto, algorithms like SAFER, IDEA, FEAL, etc fell to or were partially attacked by differential crypto. Somewhat amusingly, it appears an extended version (impossible differential cryptanalysis) wasn't known by the NSA, as their SKIPJACK algorithm was pretty well gutted following its development by Lars Knudsen and Biyam/Biryukov/Shamir.
It's the key length stuff where the NSA seems to have held things back, and at least that was public at the time.
According to the article, they're the average actual losses, not potential. If the potential losses were $4650 and the actual annual cost of FDE was $235, then you'd need to believe that every computer has a 5% chance of being stolen and exploited every year in order for FDE to be worth it; the article would have to conclude that FDE isn't worth it for the average machine in that case.
By ginning up a ludicrously overstated actual loss, though, they're able to make FDE look like a total bargain--at least until you start thinking about it enough to realize the numbers are cooked.
"The study found that the most expensive element of FDE is not the hardware or software involved, but the value of user time it takes to start up, shut down and hibernate computing systems while using FDE. "
But this study doesn't pass the smell test. Take this, for instance: "The cost savings from reduced data breach exposure was $4,650." Imagine that FDE takes the risk of data breach on a stolen disk from 100% down to 0%. And imagine that any given computer has a 1% chance each year of being stolen by someone who's going to exploit the data on it (rather than just reformat it and sell or use it). Both of those are very generous estimates.
The average value of a lost computer to my company--either in terms of profits lost or competitor's profits gained--would have to be $465,000 for the math to work. Which as a median doesn't make sense.
If it's a mean, it only makes sense because there are a handful of computers whose value is tens or hundreds of millions of dollars counterbalancing the vast array of other computers worth far less--but if that's the case, the right solution probably isn't to lump all machines together for analysis purposes, it's to segregate out the high-value targets and treat their security differently from the low-value targets.
Yep. 6-7% is typical sales tax of most of the US, though it varies from zero (in Delaware and New Hampshire) to over 11% (in parts of Arizona and until recently Illinois). And many people dodge the tax by buying online from a company that doesn't do business in their state; legally they're supposed to pay the tax themselves to their state, but it's a common one to dodge.
And I think you are mistaken on the size of some of those items. Doom came on three(?) floppy discs. Or are you expecting me to believe the geniuses who created Doom would distribute it in a format larger than the one used to fully express the editable data for the artwork and levels?
It's not uncommon for executables to be bigger than source code. And Doom distributed the runtimes that come with the compiler in the binary distro.
And Doom3's art assets are certainly larger than 9.5MB.
You're right, the images directory turns out to be an empty skeleton. My bad there. I still think the main point (that 1GB is orders of magnitude bigger than most Android games) is valid, though.
Most source code isn't measured in anything approaching gigabytes. The entire linux kernel (which is orders of magnitude bigger than any Android game source I've seen) is an 80MB archive.
Harvard Mk 1 paper tape was about 2.5" wide and carried 30 bytes per inch. That's about 2 2/3 feet per kb of data.
I've got an Android tic tac toe app that's 62k of source code. That'd be 165' of paper tape. On regular paper, that'd fan-fold to be 1' long and about 0.7" high (at 2.5" wide). Heavier card stock might quadruple that height or something like that.
The original Doom source code--with all the art and everything--is under 400k, which'd be 1000' of paper tape or a 4" high stack at 1' long and 2.5" high. It's outdated by desktop standards but in the realm of many Android games I've seen.
The Doom 3 source code with art is 9.5MB, which would be a 9' high stack.
The whole linux kernel (80MB) would be about 40 miles of tape; on regular paper that's a 35.5' high stack at 1' long and 2.5" wide. Which is cumbersome for sure, and would take almost a full day to read in at typical paper-tape scanning speeds.
So why hasn't Oracle improved the garbage collection behavior of Java Standard Edition based on the proofs of concept presented in those articles? Or if it has, why does there remain a perception that Java's garbage collection performs poorly?
Who knows, that's hardly the only major issue with Java.
If I were inclined to put a reason on it other than laziness or ignorance: Possibly they've focused primarily on throughput at the expense of latency, believing that's a better bang for their buck for wherever they're making money selling/supporting Java? Which is plausible, if desktop Java is paying out more support to 3rd party tools and toolkits (money not going to Oracle), while server-side Java helps fuel more JVM support contracts and Oracle DB sales?
Where is the drone flown by a LISP control law program ??
Oh, yeah, and it's not a drone but there were certainly a number of military robots developed by IS Robotics and others using realtime Lisp variants in the 1990s. e.g.:
The Real-Time Dynamic Languages contract consists of three separable efforts:
1. Evolutionary Design of Complex Systems (EDCS) (CLINs 0001 and 0002). 2. A Behavioral Programming Approach to Adaptive Autonomous Control (SAFER) (CLIN 0003). 3. Adding an Active Vision Head to the M4 Robot (CLIN 0004).... The portable Common Lisp subset used to develop the software is based on previous work by iRobot on L, a Common Lisp subset that currently runs on 68000-based machines. L was carefully designed and implemented for use in small, embedded processors operating under real-time constraints; it is currently in use on a number of walking, tracked, and wheeled robots being developed by iRobot.
The memory management system for L has been redesigned, making use of recent research results by the garbage collection community, particularly in the area of real-time garbage collection.
I have used quite a few languages which use GC and even though some are quite mature now, like VW Smalltalk, all of them freeze on a regular basis, just when you don't want it to happen. I have also used and developed large-scale systems which do not have this trait, and all of them were done in C or C++ and they use explicit allocation or reference-counted memory management.
1. VW Smalltalk uses a full (non-generational) tracing collector which makes no pretenses of being low-latency, let alone realtime.
2. Refcounting is a kind of GC, not something different; mark-sweep is the other most common algorithm, but it's a mistake to refer to any one method as "garbage collection" at the expense of others.
3. If you want to look at real-word implementations, Lua and squeak are probably the two most prevalent systems that use realtime GC (squeak is only mostly there--doing weird things in destructors can make it not real-time). OCaml is 2nd-generation incremental, which is not realtime but is low-latency in the common case.
Lua added incremental GC as of version 5.1 mainly because it's seeking to be used as an extension language in games which have pretty rigorous soft realtime requirements.
IBM's Metronome (AKA Websphere Realtime) is a realtime JVM that's seen a fair amount of use in the real world. It uses a newer algorithm by Bacon (who's mentioned above) that's completely hard realtime and fits the constraints of the JVM. http://researcher.watson.ibm.com/researcher/view_project.php?id=174
Garbage Collection, which kills User Experience due to unpredictable freezing of the whole program
Note that this is a product of a crappy garbage collector in the Java runtime, not intrinsic to garbage collection per se--there are plenty of well-known real time GCs that allow you to set a maximum latency on the collector.
As I recall, the most powerful bombs in the us was around 9 megatons, and they where dismantled a few years ago.
The American B41s were ~25MT; the last ones were dismantled in 1976. They were never exploded, though. The biggest US explosion was 15 MT in the infamous Castle Bravo test--that bomb was only supposed to be 5MT and the test was supposed to be secret, but unexpected reactivity of one of the lithium isotopes made it go off at triple that, causing severe fallout problems and ensuring that there was no secrecy.
The book came out in 1957; the movie in 1959. Today's average Slashdotter would have been -20 years old then. So I'm not surprised most don't know OTB or were too young to be impressed by a black and white movie with little action and starring actors they don't know.
There's also a 2000 movie and a 2008 radio drama.
For the rest of us the story was simply devastating. And totally believable. Through fear, intolerance, and stupidity, everyone everywhere dies.
Yeah. I'm shocked that it hasn't completely dominated this conversation; it's about as depressing a scenario as imaginable, and popular enough to have spawned multiple movies and be in the public conscious far beyond a strict sci-fi fanbase.
No it isn't when kids are graduating with staggering debt, taking extra courses that do NOT DIRECTLY affect my job prospects is just plain stupid!
That's like saying "how many people have staggering debt because they dropped a lot of money on a yacht that didn't directly affect their job prospects, it's just plain stupid!"--the solution isn't "figure out a way to make buying expensive boats help you earn money", it's "don't buy expensive boats if you can't afford them". A liberal arts education isn't a professional education and isn't supposed to be focused on job prospects, and if it's going to cause you staggering debt you probably ought to look at alternatives: among tons of other choices, there are job-focused programs at 4-year schools like engineering, architecture, or computer science programs; advanced occupational programs like med school and law school; college-equivalent professional schools a la nursing schools; and alternatives like automotive mechanic and other voc-tech options.
Shoehorning occupational crap into a liberal arts program is a bad idea because a) you're choosing the wrong tool for the job--it's "everything is a hammer" syndrome, and you're always going to get worse job training trying to turn an anthropology degree into occupational readiness training than you would actually going to a job-focused program; and b) you wind up ruining the liberal arts education, the purpose of which is to get a well-rounded base and learn about a particular topic in depth, notto prepare you for the workplace.
If your goal is job training, go to a job-focused program. But don't try to turn all kinds of education into job training.
(The original article, fwiw, is about algebra which is covered years before university for most folks and seems misguided mainly by dint of being out of touch with high school curriculae)
No, I don't mean that, but when you get to uni and have chosen a field to specialise in, then you should be able to do so.
Sure, and I wouldn't argue that everyone should take differential equations. But we're talking about algebra here. It's freshman in high school material, maybe sophomore if you're a bit slower than normal at math and 8th grade if you're on a normal fast-track in math (there were several kids in my year who were even another year ahead of that, but they were clearly very gifted in the arena).
I just went back and checked, and the standard track at my (normal public) high school for people pursuing an Arts/Humanities BS program after graduation has them take Algebra I (freshman), Geometry (sophomore), Algebra II (junior) and then one of Pre-Calc, Logs & Trig, and Honors Geometry as a senior. If you're a fast math student it's not unheard of to be several years ahead of that and finish Calculus I
You could easily rearrange that to be Algebra I, Geometry, "Life Math" or whatever it's called, and Algebra II to fit in a year of statistics and such (which I'm not opposed to as an idea). It'd work without sacrificing basic algebra, and for people who are a bit slower math you'd still cover the basics even if things slipped back a year.
Slashdot Mirror
Are you saying that in America you can get a criminal record for littering?
It's up to the state, but certainly you can in some states (maybe all states, I don't know). Certainly it's on your record in California and Texas, to name a couple big states at opposite ends of the political spectrum.
Not every crime is put on your record; the so-called "infractions", which are less severe than misdemeanors or felonies, are sometimes omitted. Not always, though--speeding tickets are often (lower-than-misdemeanor) infractions, but they're certainly tracked for at least a few years in most areas.
You get a lot more carryover heat gain than you'd expect.
Alton Brown's show this year he's revised in a couple of ways:
1. No breast-plate foil any more
2. Pull the bird out when the breast reaches 151F.
He did that with a 12 lb bird--the breast was 151F when he pulled it out, and after 1/2 hour of rest the carryover had raised the breast temp to 164F.
An in other areas of the U.S. they have moved enforcement from criminal to civil. Here in RI you can have up to an ounce, you can't smoke in public, and they actually license growers. It's all very quiet of course but it exists.
If you are caught in public it's like a $100 fine on the civil side, not criminal.
I hate the term "decriminalization" and wish it would go away, because it leads to confusion like this.
In RI and MA, the penalty for marijuana possession is now only a fine (no jail time or arrest). But it's still a criminal fine, not a civil fine. Criminal fines are punitive; civil fines are purely restitution for actual damages done to the government. If you dump toxic waste in a national park, you might be assessed a civil fine to pay for the costs of cleanup (on top of any possible criminal penalties for trespass, endangering people's health, etc). If you litter or speed, you can be assessed a criminal fine. Even though in common parlance people use "criminal" to mean misdemeanors and felonies, strictly speaking infractions like littering and speeding are still (very minor) crimes, and it's because of that misuse that the misleading term "decriminalization" came into play.
In general, a good way to differentiate "civil" cases (torts) vs. "criminal" cases (crimes) is to look at who the non-defendant is in the case: If you fail to fulfill the terms of a contract, or file for divorce, or are sued for child support, then the other party will hire a private attorney and sue you in civil court. If you assault someone, rape someone, commit DUI, etc, then the state's attorney (not some attorney hired by the victim) will show up and take you to a criminal court. There are many levels of criminal courts, ranging from traffic courts up to the US Supreme Court, alongside the civil court system.
The case of a "civil fine" is more complicated, but it's essentially the case where the government is showing up not to protect the general public or a third party, but to recover damage to government property. It's similar to a true civil case in that the wronged party is representing themselves in court, but in practice that's tough to distinguish because the wronged party is the government.
Yeah, the leading question is a little annoying--it should be "does anything keep you from using Linux, and if so what?" My answer would be the same as yours--I've been running Linux exclusively for 18 years now. For others, there are still applications--games or otherwise--or usability issues, driver support, or whatever that have stopped them from adopting Linux.
It's the same as the annoying "Will this be the year of the Linux desktop?" questions--the closest thing we'll ever see to a "year of the Linux desktop" was around 2000 or 2001, when you started seeing Dell and others selling Linux machines, could actually use a graphical "click here" widget to set up networking instead of needing to manually enter your gateway and network addresses by hand, didn't need to manually enter your monitor's timings as XF86 modelines, etc.
That's a far cry from saying that it was usable on the desktop by everyone, but it's basically the closest thing to a bright line we're ever liable to see, and is the point where it crossed over from "power users can use a Linux desktop if true gurus install it for them" to "regular users can use it if power users install it". From there on out, it's incremental support for apps here and there, UI improvements, driver support, etc--but there will probably never be another year of the Linux desktop as big as what we already saw 10+ years ago, it'll be slow adoption in fits and starts here and there as people's particular needs are met. The biggest global hurdles are long since passed, and we're into an (equally important, but harder to solve at one whack) forest of individual issues holding things back for people.
You're better off running your own CA and distributing that CA's public key to your internal apps. Then you can ignore outside CAs but still avoid MITM attacks.
More accurately, they _might_ be in trouble if Canadian Courts find that either applies. It'll then depend on the ongoing mess of sovereignty issues with the Council of the Haida Nations--the Haida never signed any treaties with the Canadian government, so they have a much greater claim to sovereignty than many of the First Nations and may have a legal argument that they are not signatory to those treaties (despite Canada being a signatory).
The hardest thing mixed in there was the a.out->ELF migration, for which I rebuilt everything on the system by hand sort of like a primitive LFS. It was worth doing once--you learn a fair bit in the process--but made it so I have little interest in LFS or similar distros going forward.
To be fair, the NSA don't seem to have caused problems with the S-Boxes and differential analysis doesn't seem to have worked too well
In fact, the NSA's changes to the S-boxes made DES stronger against differential cryptanalysis; it appears that they and IBM knew about diff crypto back in the 1970s and designed an algorithm to resist it even though the technique wouldn't be widely known for another 15-20 years.
Diffential crypto only "doesn't seem to have worked out so well" because it's known and algorithms are designed to withstand it--prior to the public knowledge of diff crypto, algorithms like SAFER, IDEA, FEAL, etc fell to or were partially attacked by differential crypto. Somewhat amusingly, it appears an extended version (impossible differential cryptanalysis) wasn't known by the NSA, as their SKIPJACK algorithm was pretty well gutted following its development by Lars Knudsen and Biyam/Biryukov/Shamir.
It's the key length stuff where the NSA seems to have held things back, and at least that was public at the time.
and where would mount, say, an off-the-shelf lavalier
Why not come full circle? Before it was coöpted to mean a type of microphone, lavalier originally meant a pendant that hangs from a necklace.
According to the article, they're the average actual losses, not potential. If the potential losses were $4650 and the actual annual cost of FDE was $235, then you'd need to believe that every computer has a 5% chance of being stolen and exploited every year in order for FDE to be worth it; the article would have to conclude that FDE isn't worth it for the average machine in that case.
By ginning up a ludicrously overstated actual loss, though, they're able to make FDE look like a total bargain--at least until you start thinking about it enough to realize the numbers are cooked.
At least partially:
"The study found that the most expensive element of FDE is not the hardware or software involved, but the value of user time it takes to start up, shut down and hibernate computing systems while using FDE. "
But this study doesn't pass the smell test. Take this, for instance: "The cost savings from reduced data breach exposure was $4,650." Imagine that FDE takes the risk of data breach on a stolen disk from 100% down to 0%. And imagine that any given computer has a 1% chance each year of being stolen by someone who's going to exploit the data on it (rather than just reformat it and sell or use it). Both of those are very generous estimates.
The average value of a lost computer to my company--either in terms of profits lost or competitor's profits gained--would have to be $465,000 for the math to work. Which as a median doesn't make sense.
If it's a mean, it only makes sense because there are a handful of computers whose value is tens or hundreds of millions of dollars counterbalancing the vast array of other computers worth far less--but if that's the case, the right solution probably isn't to lump all machines together for analysis purposes, it's to segregate out the high-value targets and treat their security differently from the low-value targets.
More like wine, whiskey, beer, cider, scotch, tequila, rum, and everything else: NASA was 4% of the federal budget at its peak.
Yep. 6-7% is typical sales tax of most of the US, though it varies from zero (in Delaware and New Hampshire) to over 11% (in parts of Arizona and until recently Illinois). And many people dodge the tax by buying online from a company that doesn't do business in their state; legally they're supposed to pay the tax themselves to their state, but it's a common one to dodge.
And I think you are mistaken on the size of some of those items. Doom came on three(?) floppy discs. Or are you expecting me to believe the geniuses who created Doom would distribute it in a format larger than the one used to fully express the editable data for the artwork and levels?
It's not uncommon for executables to be bigger than source code. And Doom distributed the runtimes that come with the compiler in the binary distro.
And Doom3's art assets are certainly larger than 9.5MB.
You're right, the images directory turns out to be an empty skeleton. My bad there. I still think the main point (that 1GB is orders of magnitude bigger than most Android games) is valid, though.
Most source code isn't measured in anything approaching gigabytes. The entire linux kernel (which is orders of magnitude bigger than any Android game source I've seen) is an 80MB archive.
Harvard Mk 1 paper tape was about 2.5" wide and carried 30 bytes per inch. That's about 2 2/3 feet per kb of data.
I've got an Android tic tac toe app that's 62k of source code. That'd be 165' of paper tape. On regular paper, that'd fan-fold to be 1' long and about 0.7" high (at 2.5" wide). Heavier card stock might quadruple that height or something like that.
The original Doom source code--with all the art and everything--is under 400k, which'd be 1000' of paper tape or a 4" high stack at 1' long and 2.5" high. It's outdated by desktop standards but in the realm of many Android games I've seen.
The Doom 3 source code with art is 9.5MB, which would be a 9' high stack.
The whole linux kernel (80MB) would be about 40 miles of tape; on regular paper that's a 35.5' high stack at 1' long and 2.5" wide. Which is cumbersome for sure, and would take almost a full day to read in at typical paper-tape scanning speeds.
So why hasn't Oracle improved the garbage collection behavior of Java Standard Edition based on the proofs of concept presented in those articles? Or if it has, why does there remain a perception that Java's garbage collection performs poorly?
Who knows, that's hardly the only major issue with Java.
If I were inclined to put a reason on it other than laziness or ignorance: Possibly they've focused primarily on throughput at the expense of latency, believing that's a better bang for their buck for wherever they're making money selling/supporting Java? Which is plausible, if desktop Java is paying out more support to 3rd party tools and toolkits (money not going to Oracle), while server-side Java helps fuel more JVM support contracts and Oracle DB sales?
If so it seems a short-sighted policy.
Where is the drone flown by a LISP control law program ??
Oh, yeah, and it's not a drone but there were certainly a number of military robots developed by IS Robotics and others using realtime Lisp variants in the 1990s. e.g.:
http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA399584
I have used quite a few languages which use GC and even though some are quite mature now, like VW Smalltalk, all of them freeze on a regular basis, just when you don't want it to happen. I have also used and developed large-scale systems which do not have this trait, and all of them were done in C or C++ and they use explicit allocation or reference-counted memory management.
1. VW Smalltalk uses a full (non-generational) tracing collector which makes no pretenses of being low-latency, let alone realtime.
2. Refcounting is a kind of GC, not something different; mark-sweep is the other most common algorithm, but it's a mistake to refer to any one method as "garbage collection" at the expense of others.
3. If you want to look at real-word implementations, Lua and squeak are probably the two most prevalent systems that use realtime GC (squeak is only mostly there--doing weird things in destructors can make it not real-time). OCaml is 2nd-generation incremental, which is not realtime but is low-latency in the common case.
Lua added incremental GC as of version 5.1 mainly because it's seeking to be used as an extension language in games which have pretty rigorous soft realtime requirements.
IBM's Metronome (AKA Websphere Realtime) is a realtime JVM that's seen a fair amount of use in the real world. It uses a newer algorithm by Bacon (who's mentioned above) that's completely hard realtime and fits the constraints of the JVM. http://researcher.watson.ibm.com/researcher/view_project.php?id=174
Garbage Collection, which kills User Experience due to unpredictable freezing of the whole program
Note that this is a product of a crappy garbage collector in the Java runtime, not intrinsic to garbage collection per se--there are plenty of well-known real time GCs that allow you to set a maximum latency on the collector.
See, for instance:
http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.39.4550
http://web.media.mit.edu/~lieber/Lieberary/GC/Realtime/Realtime.html
http://portal.acm.org/citation.cfm?id=604155
As I recall, the most powerful bombs in the us was around 9 megatons, and they where dismantled a few years ago.
The American B41s were ~25MT; the last ones were dismantled in 1976. They were never exploded, though. The biggest US explosion was 15 MT in the infamous Castle Bravo test--that bomb was only supposed to be 5MT and the test was supposed to be secret, but unexpected reactivity of one of the lithium isotopes made it go off at triple that, causing severe fallout problems and ensuring that there was no secrecy.
The book came out in 1957; the movie in 1959. Today's average Slashdotter would have been -20 years old then. So I'm not surprised most don't know OTB or were too young to be impressed by a black and white movie with little action and starring actors they don't know.
There's also a 2000 movie and a 2008 radio drama.
For the rest of us the story was simply devastating. And totally believable. Through fear, intolerance, and stupidity, everyone everywhere dies.
Yep.
Yeah. I'm shocked that it hasn't completely dominated this conversation; it's about as depressing a scenario as imaginable, and popular enough to have spawned multiple movies and be in the public conscious far beyond a strict sci-fi fanbase.
No it isn't when kids are graduating with staggering debt, taking extra courses that do NOT DIRECTLY affect my job prospects is just plain stupid!
That's like saying "how many people have staggering debt because they dropped a lot of money on a yacht that didn't directly affect their job prospects, it's just plain stupid!"--the solution isn't "figure out a way to make buying expensive boats help you earn money", it's "don't buy expensive boats if you can't afford them". A liberal arts education isn't a professional education and isn't supposed to be focused on job prospects, and if it's going to cause you staggering debt you probably ought to look at alternatives: among tons of other choices, there are job-focused programs at 4-year schools like engineering, architecture, or computer science programs; advanced occupational programs like med school and law school; college-equivalent professional schools a la nursing schools; and alternatives like automotive mechanic and other voc-tech options.
Shoehorning occupational crap into a liberal arts program is a bad idea because a) you're choosing the wrong tool for the job--it's "everything is a hammer" syndrome, and you're always going to get worse job training trying to turn an anthropology degree into occupational readiness training than you would actually going to a job-focused program; and b) you wind up ruining the liberal arts education, the purpose of which is to get a well-rounded base and learn about a particular topic in depth, notto prepare you for the workplace.
If your goal is job training, go to a job-focused program. But don't try to turn all kinds of education into job training.
I think we're in violent agreement.
(The original article, fwiw, is about algebra which is covered years before university for most folks and seems misguided mainly by dint of being out of touch with high school curriculae)
No, I don't mean that, but when you get to uni and have chosen a field to specialise in, then you should be able to do so.
Sure, and I wouldn't argue that everyone should take differential equations. But we're talking about algebra here. It's freshman in high school material, maybe sophomore if you're a bit slower than normal at math and 8th grade if you're on a normal fast-track in math (there were several kids in my year who were even another year ahead of that, but they were clearly very gifted in the arena).
I just went back and checked, and the standard track at my (normal public) high school for people pursuing an Arts/Humanities BS program after graduation has them take Algebra I (freshman), Geometry (sophomore), Algebra II (junior) and then one of Pre-Calc, Logs & Trig, and Honors Geometry as a senior. If you're a fast math student it's not unheard of to be several years ahead of that and finish Calculus I
You could easily rearrange that to be Algebra I, Geometry, "Life Math" or whatever it's called, and Algebra II to fit in a year of statistics and such (which I'm not opposed to as an idea). It'd work without sacrificing basic algebra, and for people who are a bit slower math you'd still cover the basics even if things slipped back a year.