I think they mean: every rocket program (since Goddard fastbaked the first potato during his first liquid fueled rocket experiments) has had weight problems.
And how does public speculation without even attempting to contact the parties fit in? In a normal newspaper, if you are going to make some claim "NASA rocket doomed", normally you give the party some chance to reply "We contacted NASA and they stated...".
Also, I'd be a bit surprised if extra weight would "doom" anything built on a modern solid rocket. It's not like there are any hard limits you run up against, its just a matter of scale and balance. This is because solid rockets are far more powerful than liquid rockets (The SRBs, for example, are the highest thrust rockets ever). Where you might run into unsolvable problems is for things like SSTO designs.
From the studies I've seen, Linux admins cost more, but less per machine. A Linux admin costs around 50% more, but could manage 2x or more in terms of machines. Governments regularly screw up IT projects, and there are numerous ways they could fail in this one, such as: (1) Retraining existing Windows admins, not hiring Linux ones (Common for a gov't job, admin has no motivation for success) (2) Hiring lots of cheap admins for Linux (Works badly for Windows, but functional. Doesn't work for Linux) (3) Too many consultants, since they don't know what they are doing (Consultants often find a way to plug their pet products) (4) User resistance, esp influential ones (If the boss can't find solitaire, heads will roll)
It gets better. The attack requires that the two processes are running on the same core with hyperthreading enabled (i.e. ALU-poor CMP). The "spy" process will be sucking up 100% cpu pretty much continuously. They also simplified the multiplication routine from OpenSSL. Even if you are running such a setup on a P4 with HT turned on (even though its often useless), and you need to run secure processes along with unsecure ones (generally not a good idea anyway), patches already exist for Linux and BSDs to address this. The patches modify the scheduler to prevent processes from different users from running on the same physical core. A half-hearted attempt is made in the paper to say that these attacks to generalize to something remote, but no details are given as to how their attack would compensate for the 100,000 fold decrease in timing accuracy to pull off the attack on even a local LAN.
Essentially they took a very impractical attack with an unlikely scenario, and created a somewhat practical attack with an unlikely scenario. Avoid the problem scenario which was raised in the prior work last year, and you are still golden.
Who, exactly, are you referring to? I don't see a single post that agrees with the article. Numerous posts disagreeing with TFA are written by people who are either from developing nations, or have spent time in one. I think you are barking up the wrong tree.
Yeah I don't know what he's talking about. I would love to have a Cell for our RoboCup robots. Computer vision algorithms will map quite nicely onto SPUs, and will allow many techniques to run at video frame rates which are currently out of the question. As far as game-ai type stuff, path planning could use denser meshes for more realistic motion, and there is even enough CPU to do some local kinematic planning as well (imagine AIs that don't drive wheeled vehicles into walls constantly). The small local memories might become an issue without some form of paging though. There should be enough CPU for better line-of-sight evaluations, allowing more stealthy AIs that consider more points they could position themselves at. Finally, there should be enough CPU for just about any coordination method you'd like to use for teams, allowing better breakup of subtasks for the AIs on a team.
Now, whether or not game developers will use the extra CPU for such things is another question altogether...
Actually I got them from the OMB's report on the 2003 budget. Just because the "mandatory" parts of our budget require changes in law to change them, doesn't mean you can just cut out 2/3 of the budget and not even consider where the money goes. I can use your numbers, but I'm going to use the big picture. Medicare and Medicaid do count, even though you pretend they don't exist ($670+ B), since they are defined by coverage and not cost, and costs are spiraling out of control while the government tacitly accepts the "market" price (average 15% increase per year for those programs).
The 2007 budget graph conflates the military spending with the more-or-less direct cost of the Iraq war. I agree that the Iraq war is a true and total waste of money; You'll notice I mention it in the GP as a waste. That wasn't what GGP said though, they referred to "the military". That's what I disagree with; I think the 17-18% spent to maintain our military is fine, as it is a low fraction of our GDP (4%), and ranks us 26th in the world. The growth of 7% per year should preferably be a lower, but compounds a lot slower than 15% (5 versus 11 years to double). If the Dems get us out of Iraq in some timely fashion, I don't expect that 7% increase to last that long either. When it comes to total health costs, however, we're #1 in the world as %GDP, and we don't even have the best care. Thus as you might expect, the total for all the social programs is going nuts, and is not going to be maintainable.
In short (using your source for numbers, including the war), giving the military a day off would save $1.7B, so the GGP was off by a factor of 9 regarding NASA's budget. After the war, it will be a factor of 11 off. I think it is a common misconception, so I was just calling that out, that's all. Say and think whatever you want.
The normal delay between plays is a maximum of 40 seconds. Each team also gets three 30 second timeouts each half. This is not that bad compared to tennis or baseball. However, additional timeouts have been added to the modern NFL game to allow for commercials, extending regular timeouts to up to two minutes, and adding one-minute breaks after changes of possession. These suck, but that's where the money comes from. Canadian football and College football do not have most of these extra breaks, leading to a faster paced game. Of course, given the existence of golf and cricket, no other sports can really be called too slow or too long.
Back on topic, the American football team "Seattle Seahawks" is owned by a MSFT founder. Buying support therefore comes naturally to them, so the Thai IT minister's newfound support shouldn't be a surprise.
P.S. It would be funny if Steve Jobs bought one of the San Francisco Bay area teams.
Give me a break; The term "British English" doesn't mean that no other language can be spoken in Britain, just like saying "Swiss German" doesn't mean that other Swiss don't speak a dialect of French. Every language and dialect is some derivative and/or combination of others, so its hard enough to refer to them in any universal way, let alone with someone being pedantic about it. The ISO language code for the dialect is currently "en-gb". Get them to officially change it and I'll refer to it however you want. How about the Queen's English, I look forward to generating an "en-queen" locale.
It's British English you idiot. Try "checkered" versus "chequered" in Google, and you'll get 4.6 times as many hits for the British spelling. Yes, there are actually a bunch of people out there who don't speak American English like we do. And 99% of the time, you will have absolutely no trouble guessing which word they meant. Now, if they could just return the favor when I use the word "soccer" and not try to "correct" me, we could live a happy tolerant world.
Or you could make a tiny dent in US health and social programs, which have ballooned to 50% of the US budget (not including health research money which makes up a big chunk of the 20% discretionary funding). The military budget is 17%, which includes a lot of basic research (which US companies won't invest in) and general use things such as GPS. While we could already have a nice ISS and lunar output if we had forgone the Iraq war, by 2020 we'll be broke even if defense spending was zero, since health care will eat up everything else (and we don't even have universal public health care). It's amazing that we're willing to spend so much to add a year or two to our lives, while possibly hurting the long-term survival of humanity as a whole.
The Nature Conservancy is a really nice organization that works that way. A friend who donates to it pointed it out to me. In short, they use donations to buy at-risk land, and also get conservation easements from private land owners who wish to protect their own land. Advocacy organizations can spend lots of money lobbying, and potentially fail anyway, while its extremely difficult to fight someone when they actually own the land.
Yep you are right. The stats I dug up a while ago were incorrect. Maybe it only counted deployed troops (which would still be incorrect). I wouldn't count reserves however, as some countries maintain practically their entire male population as reserves (Switzerland, Israel, Taiwan, Singapore). The US reserves aren't that large in comparison.
There's also reality and fantasy. The reality is the most heavily defended and hostile border in the world, and that is not going to change immediately.
and ample justification for leaving things around to kill any kid that wanders past in fifty years time?
A sentry gun uses electricity. Presumably you would turn it off when its no longer needed. It's not going to run for 50 years unattended, and probably will only be set to autonomous operation during a crisis. Please tell me how this is not an improvement over land mines in terms of long-term safety?
It may not be in your backyard now
The DMZ is not a backyard.
but remember when the rules and morality are discarded there is little to stop it being in your backyard.
Morality was discarded because you discarded the rest of my post, which dealt with that. Note for example, the entire second paragraph of my post, which expresses optimism we can maybe use this to expand the land-mine banning treaty.
We have enough problems now from people building on old artillery ranges
Machine gun bullets are not explosive. It would be no more dangerous than an old US civil war battlefield.
how do you think it will be when you have places that have been mined to prevent the threat of terrorism on your own soil get used for different purposes?
I thought you claimed to be talking about the "similar problems" that a sentry gun has compared to a land mine? Why would it be the least bit difficult to turn off and remove a sentry gun?
Reality dictates defense of hostile borders. As I pointed out in the original post: Would it be nice to live in a world where such things were not needed? Of course. Wish as we might, the reality is that we don't (yet) live in that world. So, if the choice is between land mines and an automated sentry gun, the latter has many advantages. Feel free to disagree.
You took something out of context, and replied only to that. You also applied morality to a sentence which was simply noting a reality, not passing a judgement either way. It's like me saying "Murder is a fact of life that modern nations have to deal with.", to which you reply "Murder is wrong." No kidding, we don't necessarily disagree, and that's not the point. THE ENTIRE POINT OF MY POST WAS THAT NOW WE HAVE NO EXCUSE FOR LANDMINES.
Apples and oranges. Take that Windows app, and replace some of the DLLs in its program folder with the newest versions, and see how well it runs. Just about everything on Windows is effectively statically linked, save for a few core libraries. If I statically link a Linux app, I don't have any library issues either (this is what many commercial closed-source apps on Linux do).
Of course, some "important" business apps dive into all sorts of details that they have no business accessing, and feel the need to reimplement the OS whenever possible *cough*Oracle*cough*. They will of course break on an upgrade, as they were effectively designed to do so. Keeps the support contracts up to date I guess.
That's because Windows libraries never get past 1.x. Features/cruft get added, but the original design decisions can never be altered, even if they turn out to be terrible. No thanks.
There are a lot of libraries in Linux/*nix, so you must choose wisely before you commit to one. You have to make sure that their design goals overlap with your needs. If you want stability in interfaces, only choose libraries which are mature, and developed by a community that cares about interface stability.
A driver is hardly the same as an application. Linux has never claimed to guarantee stability for the driver interface, and in fact steadfastly refuses to do so, so that improvements can be made. Linux goes to great pains however to maintain the application interface (ABI). I can run plenty of old applications, in particular many X apps work just fine. System tools may break, but they are diving into internals that aren't really part of the official interface. This is no different in Windows, where disk and AV tools will need to be upgraded with Windows. Show me some hardware for which there is a single driver which works for Win98, Win2k, and Vista... Drivers!=Applications. Even then, I rarely have issues, as I run exactly one driver (NVidia graphics) which doesn't come distributed with Linux itself. Every other driver upgrade is handled by the kernel developers, and waiting a bit after releases to look for reviews of breakage. Use a distribution kernel if you want stability, and then you will almost never have any issues at all.
Well.. and the most probable response of said madman will be to speed up the nuclear bomb program,
If you haven't noticed, that is his response to anything. If we tell him "we're not going to give you oil and grain for free anymore," he considers it an act of war. If we tell him "we're not going to sell you things that you can turn into weapons," he also considers it an act of war. Should we instead try appeasement? We tried that in 1994, and he repaid us by continuing with his bomb research secretly anyway. NK knows it is screwed in the long run, but they also know that they have won in the short term; We have no good options, and he will not stop his research for anything.
And not to mention that the real loosers in this solution are the people who are trying to escape from north korea.
First of all, it's LOSERS. Second, very, very few NK citizens escape across that border, as they will get shot by NK soldiers. It's just like the East/West German border. NK has electric fences and machine guns. You get shot if you go swimming on the beach anywhere near the border. Also, there are huge minefields from both sides, and armed human guards on the SK side as well. The only way you can escape is in the northern border with China... even then, it is quite dangerous and many do not make it. In this documentary, they talk about crossing the northern border. Imagine something 10x as dangerous, and that is what the southern border is like.
So, the sentry gun will not make it any more dangerous than it already is, and it at least gives a human operator a chance to tell a trespasser to halt. You don't get that option with an electric fence or land mines.
How about you read the rest of my post before commenting? A usable sentry gun means we can decrease or end entirely the use of land mines for hostile border defense.
Think about the alternatives... Given the 1+ million strong army north of the border, and the questionable sanity of the leader controlling it, that border must be defended. The numerical superiority means some defenses must be automated, leaving land mines as the only existing technology. This robot is far better than a land mine however; It can be switched off, can be configured to give a warning, and can be removed easily when it is no longer needed. Land mines have none of these properties.
Would it be nice to live in a world where such things were not needed? Of course. I'm not going to blame the South Koreans at all though, given the realities of their situation. Maybe it will even let more countries sign the land mine treaty/ban. The US, for example, could buy these for defending Guantanamo, and remove the land mines we have placed there.
Slashdot Mirror
I think they mean: every rocket program (since Goddard fastbaked the first potato during his first liquid fueled rocket experiments) has had weight problems.
And how does public speculation without even attempting to contact the parties fit in? In a normal newspaper, if you are going to make some claim "NASA rocket doomed", normally you give the party some chance to reply "We contacted NASA and they stated...".
Also, I'd be a bit surprised if extra weight would "doom" anything built on a modern solid rocket. It's not like there are any hard limits you run up against, its just a matter of scale and balance. This is because solid rockets are far more powerful than liquid rockets (The SRBs, for example, are the highest thrust rockets ever). Where you might run into unsolvable problems is for things like SSTO designs.
From the studies I've seen, Linux admins cost more, but less per machine. A Linux admin costs around 50% more, but could manage 2x or more in terms of machines. Governments regularly screw up IT projects, and there are numerous ways they could fail in this one, such as:
(1) Retraining existing Windows admins, not hiring Linux ones (Common for a gov't job, admin has no motivation for success)
(2) Hiring lots of cheap admins for Linux (Works badly for Windows, but functional. Doesn't work for Linux)
(3) Too many consultants, since they don't know what they are doing (Consultants often find a way to plug their pet products)
(4) User resistance, esp influential ones (If the boss can't find solitaire, heads will roll)
Think of the ugliest possible art stretched into three dimensions
:)
I thought we were talking about SL, not WoW.
It gets better. The attack requires that the two processes are running on the same core with hyperthreading enabled (i.e. ALU-poor CMP). The "spy" process will be sucking up 100% cpu pretty much continuously. They also simplified the multiplication routine from OpenSSL. Even if you are running such a setup on a P4 with HT turned on (even though its often useless), and you need to run secure processes along with unsecure ones (generally not a good idea anyway), patches already exist for Linux and BSDs to address this. The patches modify the scheduler to prevent processes from different users from running on the same physical core. A half-hearted attempt is made in the paper to say that these attacks to generalize to something remote, but no details are given as to how their attack would compensate for the 100,000 fold decrease in timing accuracy to pull off the attack on even a local LAN.
Essentially they took a very impractical attack with an unlikely scenario, and created a somewhat practical attack with an unlikely scenario. Avoid the problem scenario which was raised in the prior work last year, and you are still golden.
Who, exactly, are you referring to? I don't see a single post that agrees with the article. Numerous posts disagreeing with TFA are written by people who are either from developing nations, or have spent time in one. I think you are barking up the wrong tree.
Yeah I don't know what he's talking about. I would love to have a Cell for our RoboCup robots. Computer vision algorithms will map quite nicely onto SPUs, and will allow many techniques to run at video frame rates which are currently out of the question. As far as game-ai type stuff, path planning could use denser meshes for more realistic motion, and there is even enough CPU to do some local kinematic planning as well (imagine AIs that don't drive wheeled vehicles into walls constantly). The small local memories might become an issue without some form of paging though. There should be enough CPU for better line-of-sight evaluations, allowing more stealthy AIs that consider more points they could position themselves at. Finally, there should be enough CPU for just about any coordination method you'd like to use for teams, allowing better breakup of subtasks for the AIs on a team.
Now, whether or not game developers will use the extra CPU for such things is another question altogether...
Actually I got them from the OMB's report on the 2003 budget. Just because the "mandatory" parts of our budget require changes in law to change them, doesn't mean you can just cut out 2/3 of the budget and not even consider where the money goes. I can use your numbers, but I'm going to use the big picture. Medicare and Medicaid do count, even though you pretend they don't exist ($670+ B), since they are defined by coverage and not cost, and costs are spiraling out of control while the government tacitly accepts the "market" price (average 15% increase per year for those programs).
The 2007 budget graph conflates the military spending with the more-or-less direct cost of the Iraq war. I agree that the Iraq war is a true and total waste of money; You'll notice I mention it in the GP as a waste. That wasn't what GGP said though, they referred to "the military". That's what I disagree with; I think the 17-18% spent to maintain our military is fine, as it is a low fraction of our GDP (4%), and ranks us 26th in the world. The growth of 7% per year should preferably be a lower, but compounds a lot slower than 15% (5 versus 11 years to double). If the Dems get us out of Iraq in some timely fashion, I don't expect that 7% increase to last that long either. When it comes to total health costs, however, we're #1 in the world as %GDP, and we don't even have the best care. Thus as you might expect, the total for all the social programs is going nuts, and is not going to be maintainable.
In short (using your source for numbers, including the war), giving the military a day off would save $1.7B, so the GGP was off by a factor of 9 regarding NASA's budget. After the war, it will be a factor of 11 off. I think it is a common misconception, so I was just calling that out, that's all. Say and think whatever you want.
The normal delay between plays is a maximum of 40 seconds. Each team also gets three 30 second timeouts each half. This is not that bad compared to tennis or baseball. However, additional timeouts have been added to the modern NFL game to allow for commercials, extending regular timeouts to up to two minutes, and adding one-minute breaks after changes of possession. These suck, but that's where the money comes from. Canadian football and College football do not have most of these extra breaks, leading to a faster paced game. Of course, given the existence of golf and cricket, no other sports can really be called too slow or too long.
Back on topic, the American football team "Seattle Seahawks" is owned by a MSFT founder. Buying support therefore comes naturally to them, so the Thai IT minister's newfound support shouldn't be a surprise.
P.S. It would be funny if Steve Jobs bought one of the San Francisco Bay area teams.
Give me a break; The term "British English" doesn't mean that no other language can be spoken in Britain, just like saying "Swiss German" doesn't mean that other Swiss don't speak a dialect of French. Every language and dialect is some derivative and/or combination of others, so its hard enough to refer to them in any universal way, let alone with someone being pedantic about it. The ISO language code for the dialect is currently "en-gb". Get them to officially change it and I'll refer to it however you want. How about the Queen's English, I look forward to generating an "en-queen" locale.
IN SOVIET RUSSIA, HTML tags you!
e. e. cummings likes lowercase to avoid the lame(ness) filter.
It's British English you idiot. Try "checkered" versus "chequered" in Google, and you'll get 4.6 times as many hits for the British spelling. Yes, there are actually a bunch of people out there who don't speak American English like we do. And 99% of the time, you will have absolutely no trouble guessing which word they meant. Now, if they could just return the favor when I use the word "soccer" and not try to "correct" me, we could live a happy tolerant world.
(1) probably
(2) yes it will be sad
Or you could make a tiny dent in US health and social programs, which have ballooned to 50% of the US budget (not including health research money which makes up a big chunk of the 20% discretionary funding). The military budget is 17%, which includes a lot of basic research (which US companies won't invest in) and general use things such as GPS. While we could already have a nice ISS and lunar output if we had forgone the Iraq war, by 2020 we'll be broke even if defense spending was zero, since health care will eat up everything else (and we don't even have universal public health care). It's amazing that we're willing to spend so much to add a year or two to our lives, while possibly hurting the long-term survival of humanity as a whole.
The Nature Conservancy is a really nice organization that works that way. A friend who donates to it pointed it out to me. In short, they use donations to buy at-risk land, and also get conservation easements from private land owners who wish to protect their own land. Advocacy organizations can spend lots of money lobbying, and potentially fail anyway, while its extremely difficult to fight someone when they actually own the land.
Yep you are right. The stats I dug up a while ago were incorrect. Maybe it only counted deployed troops (which would still be incorrect). I wouldn't count reserves however, as some countries maintain practically their entire male population as reserves (Switzerland, Israel, Taiwan, Singapore). The US reserves aren't that large in comparison.
Ok, then let's look at it point by point.
So there is no good or evil - just shades of grey
There's also reality and fantasy. The reality is the most heavily defended and hostile border in the world, and that is not going to change immediately.
and ample justification for leaving things around to kill any kid that wanders past in fifty years time?
A sentry gun uses electricity. Presumably you would turn it off when its no longer needed. It's not going to run for 50 years unattended, and probably will only be set to autonomous operation during a crisis. Please tell me how this is not an improvement over land mines in terms of long-term safety?
It may not be in your backyard now
The DMZ is not a backyard.
but remember when the rules and morality are discarded there is little to stop it being in your backyard.
Morality was discarded because you discarded the rest of my post, which dealt with that. Note for example, the entire second paragraph of my post, which expresses optimism we can maybe use this to expand the land-mine banning treaty.
We have enough problems now from people building on old artillery ranges
Machine gun bullets are not explosive. It would be no more dangerous than an old US civil war battlefield.
how do you think it will be when you have places that have been mined to prevent the threat of terrorism on your own soil get used for different purposes?
I thought you claimed to be talking about the "similar problems" that a sentry gun has compared to a land mine? Why would it be the least bit difficult to turn off and remove a sentry gun?
Reality dictates defense of hostile borders. As I pointed out in the original post: Would it be nice to live in a world where such things were not needed? Of course. Wish as we might, the reality is that we don't (yet) live in that world. So, if the choice is between land mines and an automated sentry gun, the latter has many advantages. Feel free to disagree.
You took something out of context, and replied only to that. You also applied morality to a sentence which was simply noting a reality, not passing a judgement either way. It's like me saying "Murder is a fact of life that modern nations have to deal with.", to which you reply "Murder is wrong." No kidding, we don't necessarily disagree, and that's not the point. THE ENTIRE POINT OF MY POST WAS THAT NOW WE HAVE NO EXCUSE FOR LANDMINES.
Apples and oranges. Take that Windows app, and replace some of the DLLs in its program folder with the newest versions, and see how well it runs. Just about everything on Windows is effectively statically linked, save for a few core libraries. If I statically link a Linux app, I don't have any library issues either (this is what many commercial closed-source apps on Linux do).
Of course, some "important" business apps dive into all sorts of details that they have no business accessing, and feel the need to reimplement the OS whenever possible *cough*Oracle*cough*. They will of course break on an upgrade, as they were effectively designed to do so. Keeps the support contracts up to date I guess.
That's because Windows libraries never get past 1.x. Features/cruft get added, but the original design decisions can never be altered, even if they turn out to be terrible. No thanks.
There are a lot of libraries in Linux/*nix, so you must choose wisely before you commit to one. You have to make sure that their design goals overlap with your needs. If you want stability in interfaces, only choose libraries which are mature, and developed by a community that cares about interface stability.
A driver is hardly the same as an application. Linux has never claimed to guarantee stability for the driver interface, and in fact steadfastly refuses to do so, so that improvements can be made. Linux goes to great pains however to maintain the application interface (ABI). I can run plenty of old applications, in particular many X apps work just fine. System tools may break, but they are diving into internals that aren't really part of the official interface. This is no different in Windows, where disk and AV tools will need to be upgraded with Windows. Show me some hardware for which there is a single driver which works for Win98, Win2k, and Vista... Drivers!=Applications. Even then, I rarely have issues, as I run exactly one driver (NVidia graphics) which doesn't come distributed with Linux itself. Every other driver upgrade is handled by the kernel developers, and waiting a bit after releases to look for reviews of breakage. Use a distribution kernel if you want stability, and then you will almost never have any issues at all.
Well.. and the most probable response of said madman will be to speed up the nuclear bomb program,
If you haven't noticed, that is his response to anything. If we tell him "we're not going to give you oil and grain for free anymore," he considers it an act of war. If we tell him "we're not going to sell you things that you can turn into weapons," he also considers it an act of war. Should we instead try appeasement? We tried that in 1994, and he repaid us by continuing with his bomb research secretly anyway. NK knows it is screwed in the long run, but they also know that they have won in the short term; We have no good options, and he will not stop his research for anything.
And not to mention that the real loosers in this solution are the people who are trying to escape from north korea.
First of all, it's LOSERS. Second, very, very few NK citizens escape across that border, as they will get shot by NK soldiers. It's just like the East/West German border. NK has electric fences and machine guns. You get shot if you go swimming on the beach anywhere near the border. Also, there are huge minefields from both sides, and armed human guards on the SK side as well. The only way you can escape is in the northern border with China... even then, it is quite dangerous and many do not make it. In this documentary, they talk about crossing the northern border. Imagine something 10x as dangerous, and that is what the southern border is like.
So, the sentry gun will not make it any more dangerous than it already is, and it at least gives a human operator a chance to tell a trespasser to halt. You don't get that option with an electric fence or land mines.
How about you read the rest of my post before commenting? A usable sentry gun means we can decrease or end entirely the use of land mines for hostile border defense.
The US army is about 1/5 that size in terms of active troops.
Think about the alternatives... Given the 1+ million strong army north of the border, and the questionable sanity of the leader controlling it, that border must be defended. The numerical superiority means some defenses must be automated, leaving land mines as the only existing technology. This robot is far better than a land mine however; It can be switched off, can be configured to give a warning, and can be removed easily when it is no longer needed. Land mines have none of these properties.
Would it be nice to live in a world where such things were not needed? Of course. I'm not going to blame the South Koreans at all though, given the realities of their situation. Maybe it will even let more countries sign the land mine treaty/ban. The US, for example, could buy these for defending Guantanamo, and remove the land mines we have placed there.