A company will always be somewhere in the spectrum between two extremes:
1) They have knowledgeable, competent staff in the areas of computer security, who can get all the practical computer security that's possible with minimal money spent on 3rd party products and consulting.
2) They don't have anyone who knows what they're doing about security, so they just fall into a cycle of throwing money at the problem, fail to get it right, throw more money, repeat ad nauseum. The money gets spent on consultants and on whatever whizbang buzzword laden security product the PHCIO just heard about in his favorite IT Mag for Dummies.
Hence the companies that are the most secure tend to be the ones spending the least money on security. I get the feeling that shops which are closer to category 2 are going to read the Gartner summary and decide to cut their IT security budgets in half in hopes that fixes all their problems, instead of investigating the real underlying issues: hiring competent people who can do security.
I agree, I don't see any reason for all the GoogleBacklash we're seeing recently. Google is still a fine company that employs a whole lot of really smart guys that I greatly admire, and lets them run wild doing really cool things The Right Way. I have no problem with how they're operating GMail, as a matter of fact I love it. I'm not so sure about this browser thing - I'm hoping it's just skinning and Google-related extensions on Firefox, or at least Gecko-based in general.
Cisco is the de-facto brand of networking gear for standard stuff like Ethernet. How much better are these high-performance switches people are talking about as suggestions in the comments here? This is not a rhetorical question, I just realy want to know and I'm too lazy and uninterested to look into it myself, but not lazy enough to stop typing this slashdot post. Is it enough to be worth going with non-Cisco for HPC clusters that use Ethernet-based interconnects? I know Cisco isn't infallible, but for all kinds of reasons they're a good bet in networking gear come purchasing time, at least outside this HPC cluster business.
Err.. well, I was half blind - the crypto benchmark part of the article does read that the 64-bit binaries are faster, I think, which agrees with the graphs - I just misread it after reading the conclusion first. But in the conclusion, they make the statement "Unfortunately, there were still several cases where 64-bit binaries performed slower; John the Ripper and OpenSSL being two of those examples."
Maybe I'm blind (seems to be the case), but I stared at the OpenSSL graph results, and I see the opposite of what the written text claims about them both on the benchmark page and in the conclusion. The written statements were to the effect that the 64-bit binaries sucked and that it was probably because OpenSSL was so heavily 32-bit optimized - but when you mouse-over for the 64-bit OpenSSL graphs of AES and RSA, the 64-bit binary result numbers look like they're trouncing the 32-bit counterpart binary on the same processor, as well as everything else in the test. What gives?
If they put out an ad right now offering $100k+ for any bigtime spammer (or with a spam sample saying "Get the guy who sends *these*), I would probably be among many people who would track these people from their spam all the way back to their home addresses and fly out there and physically confirm it and haul them in if we had to, for that kind of money.
Re:Stupid site not worthy of slashdot
on
Mock World Vote
·
· Score: 1
More importantly, even if all of the above issues were addressed, you're really getting a picture of how internet-connected probably engineering minded individuals sprinkled around the world that happened to catch wind of this site would have voted, which is nothing like the true population of the entire world would vote, I'm sure.
I was under the impression that this earlier UT "long nanotube" was actually made by braiding smaller nanotubes together into a larger "strand". The difference here is that the 4cm world record is for a single tube, no braiding things together.
Well, I'm not a scientist of any type, much less a physicist. I guess my layman brain just happens to be more aligned with the physicists' worldview:)
After reading your post, I'm inclined to think that the whole point of science is to construct a reductionist set of explanations that branches from the lowest levels of physics upwards to explaining everything else that needs explaining. Science is just how we conduct experiments in an attempt to flesh out this reductionist knowledge.
What happens when you farm enough wind power to replace all the fossil fuel burning that goes on in the world today? Does it suck enough life out of big wind current patterns to affect them in any significant way? (Which could, in theory, cause storms, sea level changes, etc...)
His point is that you're a crackpot. I've read your writing too. You want to know why all the other sciences deal directly with causality and physics doesn't? I'll tell you. It's because all of the other sciences are essentially specialized subfields of physics, and generally abstracted to a higher level than physicists work at. Ultimately, in any other branch of science, when you follow the scientific chain of causality, it leads back to physics, every time. Physics is the study of the lowest-level functioning of the reality we live in, whatever it is. You can only ask for a causal Why at the lowest layers of current physics (at any given time) for two reasons - either as a philosophical side-track that has little to do with furthering physics, or to prompt the unpeeling of another layer in the physics onion (Why do protons and neutrons and electrons behave as they do? Aha, it's those damn quarks, we've unravelled another onion layer, yay for physics, and so forth). Eventually we'll reach the very bottom, if one exists.
Well, perhaps my personal views are different than yours on this matter. But just as you gave the example of being able to be good freinds with someone of opposing worldview, I believe it's possible, and perhaps refreshing an interesting, to do the same in a LTR. I'm not saying one should seek that out, but I am saying that I would let chemistry and whatever else decides who ends up with who be the primary factors, and political alignment would be pretty low priority. I suppose if the worldviews were at opposite ends of the spectrum and both individuals were sycophants of their political faith, allowing it to dictate all aspects of their life, then it clearly wouldn't work, but then I would think those people are giving undue importance to such matters, as I said.
You're right of course, but I bet if you did as a block-level jit thing and kept the original asm code around in memory too, you could work it all out. Sounds like something Transmeta probably had to work through in their translating microcode I'm sure.
It's not that simple, which is why other projects like Wine haven't taken that approach. Ultimately some software ends up touching things in a hardware-specific way that the kernel does not abstract for you (it merely provides a pass-through interfaces for accessing some hardware in these cases), in which case the library doing so has to be translated as well.
Sadly enough, legal matter are the most important ones. If you value your own freedom enough that you don't want to risk ending up in federal pound-me-in-the-ass prison, then you really have only two options: Either learn the basic of the legal system regarding self-defense and deadly force, or choose to go unarmed and defenseless so that you'll be less likely to accidentally cross those laws the wrong way. And of course, if you value your own freedom, then option two really isn't an option.
I think you're probably intelligent enough to realize that there's an extremely low probability you'll actually be elected President of the United States this year. Do you think the Libertarian Party will ever win a Presidential Election? If so, what's the plan for getting there, and how long do you think it will take?
Actually, I could conceive a brilliant software engineer coming up with a universal translation mechanism that turns x86 assembler into functionally equivalent PowerPC assembler, or vice-versa, or to other platforms. I believe IBM had been funding research in thsi area for quite some years now.
What sets off the BS detector for me is the APIs. They consistently state that they can do this for any OS. You have to do API translation, and you'd have to do that per OS, and it's a staggering volume of work to get all the APIs translated (think Wine project, just trying to do windows->linux api on a single shared hardware platform). When my linux binary calls any given kernel, C library, or even other common library (readline?, pthreads?, opengl?, etc..), those calls all have to be translated to equivalent MacOS or Windows API calls.
SBC is a baby bell in heredity. I'm assuming TDS Metrocom is a CLEC, and that they are in fact using re=sold SBC lines, since I haven't heard tell of any true alternative LECs running their own physical networks. Your lines are still serviced and providiosned by SBC, subcontracted to TDS Metrocom. TDS is probably just operating a more efficient business on a lower profit margin than the front offices of SBC do.
You don't want 50 seperate startup companies all laying their own custom fiber or coax networks through your city redundantly, when you know that when it all shakes out, at best 3 will survive. Because physical infrastructure is involved, typically communities award a contract to a single player, who will provide said infrastructure - and they tend to keep that contract going because of the infrastructure (there would be financial issues switching to a new provider and possibly having to pay for them to buy the old infrastructure from the previous company).
The government's answer to these competitive problems to date in the electric and telecomm markets is to enforce a really stupid form of competition, and require incumbent telcos and power providers to share their physical network with startups (where "share" means basically resell). So now You can choose Traditional Company A, or one of A's 50 crappy resellers who have no physical infrastructure of their own.
The Right Thing, IMHO, is that municipalities (or states, or whatever) should be putting out bids and taking contracts to build physical cabling infrastructure of various types for the area, and also contracts (perhaps from the same provider, but not neccesarily) to maintain said cabling infrastructure. The cabling infrastructure is then owned by the municipality itself, and it terminates at certain wire-centers where competitive providers can hook in with their own real equipment (much like internet NAPs in some ways - a shared facility where people co-loc edge equipment).
Anyone else find it odd that since this story it's been an abnormally long time for/. to go with no new stories on the front page? Did the editors all go out drinking together or something?
I don't know who's confused about what here, but I'll run down the relevant laws right quick to clear this up:
1) NFA (1934) - Esentially "banned" full-auto weapons for civilian use, as well as a whole lot of other things (heavy duty explosives, big tank rounds, grenades, etc... you get the picture, silencers, too, I believe are covered). The reason I put "banned" in quotes is because it didn't really ban this stuff - it just required that you get a proper license from the Federal BATF for each weapon on this list you wanted to own, and pay them a tax. They made getting that license very difficult for most people, which makes it effectively, mostly, like a ban. The key point in the process of getting the license is that you have to get a local head of law enforcement (police chief, county sheriff, etc) to write in a personal letter stating that they authorize your ownership of a full-auto weapon, and they know you and love you, and think your full auto weapon will be a great asset to the community (good luck with that one, especially if you're not an upstanding member of your community to begin with - the vast majority of police chiefs wouldn't write that letter for their own officers' personal gun collections, much less your ass). Note that the license is per weapon, you have to do the whole process over again for each individual gun (or silencer, or mortar, or whatever your banned item of choice is).
2) The 1986 law (whatever it's called) - this went a little further and said that civilian ownership of any full auto manufactured post-1986 was flat out illegal, and you can't even get a license to cover it. So now, pre-1986 full-auto are ok (if you have the license, which you're pretty unlikely to have), but post-1986 full-auto guns are always illegal, no exceptions.
3) The 1994 Clinton Gun Ban (which expires tonight at midnight, thank god) - was just about completely pointless and didn't change any of the above. It has nothing to do with full-auto weapons, only semi-auto weapons which are functionally identical to many non-banned "hunting" rifles. And it was cosmetic in nature. Statistics over the past 10 years have proved that it had zero practical effect on crime or collecters' gun stockpiles of varioues categories of weapons, and it's finally going off the books.
Now as a side note, it is useful to realize that there's a loophole of sorts in the original 1934 "ban". The whole thing about getting police chief signoff does not apply to you if you happen to be a corporation properly registered and operating legally within your state. Therefore it is conceivable and doable (many have done it) to set up a personal corporation, file for the license under the corporation's name, and get your license to own those restricted weapons. The 1986 thing still applies of course, so you have to buy ones manufactured (and registered as NFA firearms in the US, I should point out) before that date - which still means no Glock 18's. Since there is a limited pool of pre-1986 full-auto weapons that are properly NFA registered in the US, and a great many gun nuts and collectors, the prices of these pre-86 full-autos is sky high to boot. You can get a cheap crappy submachine gun that was converted from semi-auto (and registered converted as NFA before 86...) for a few thousand if you're lucky, but if you want a Brand Spankin New In Box Colt M16A2 (mfg before 86, but never fired or messed with) today, expect prices upwards of 12-15k.
Yeah I should've mentioned the higher percentage stuff. I usually buy it at the pharmacy in the same section as the regular "rubbing alcohol". The purer bottles are usually labelled for use with hypodermic needles (to swab the area before you stick the needle in).
The problem is mostly a legal one, as anyone who's taken a concealed handgun course will tell you. The laws vary by state, but many are like Texas, where I'm at and can reasoanbly describe.
One factor is threat escalation. When one party initiates the use force, then the other party is justified in using force as a defense. When the first party escalates to deadly force, then the other party is justified in using deadly force in defense. Whoever initiates each escalation is the agressor who will likely be criminally convicted, whoever defends without escalating further up the chain of "nothing->force->deadly force" than the other party has already done is in the clear on defensive grounds. Where this all ties back in to the point is that while drawing a weapon and pointing it at someone only constitutes "force" (and is therefore legally no different than shouting, pushing, or grabbing their arm), firing the weapon, even into the ground or air as a warning shot, constitutes deadly force.
In Texas in particular, property rights are strong, and you can initiate deadly force in response to certain property crimes under certain condititions justifiably, even though the person committing the property crimes isn't neccesarily using deadly force against you. It's complicated, but a good rule of thumb for this stuff is that if it's dark, the guy is either inside your house and not clearly visible (lights are out), or it's dark and the guy is in your lawn showing signs of attempted arson (gas can in hand), you can shoot.
But just as one should never fire a warning shot (as it is a meaningless threat escalation and puts you on the wrong side of certain legal issues), one should also never "shoot to kill", or at least never phrase it that way to the cops who show up afterwards or the grand jury you'll be facing even in a defensive case. The important thing is that you were "shooting to stop" (either stop the property crime under the right circumstances, shooting to stop the threat on your life or that of others, shooting to stop "deadly force" actions against you or others, or any combination of the above). That also means that once the person does "stop" (dead, severely disabled and immobilized, running away, whatever), you are obligated to stop shooting, or once again you're on the wrong side of the law. Therefore a practical consideration to keep in your head (But enver say out loud) is that when you do make that shot to "stop", it better be deadly - because chances are after the initial hit the guy *will* stop one way or the other, and if he stays alive, he's likely to sue in civil court for his enduring medical problems if you winged him.
So, in summary, make sure you know your state's laws about when the use deadly force is authorized (A concealed handgun course in states that offer it is a great source of information and training) - and shoot to kill, but don't ever admit to shooting to kill, only shooting to stop - and do stop shooting when they stop aggressing. I would personally recommend a double-tap to the chest for your opening and closing volley.
Slashdot Mirror
Long live SPF
A company will always be somewhere in the spectrum between two extremes:
1) They have knowledgeable, competent staff in the areas of computer security, who can get all the practical computer security that's possible with minimal money spent on 3rd party products and consulting.
2) They don't have anyone who knows what they're doing about security, so they just fall into a cycle of throwing money at the problem, fail to get it right, throw more money, repeat ad nauseum. The money gets spent on consultants and on whatever whizbang buzzword laden security product the PHCIO just heard about in his favorite IT Mag for Dummies.
Hence the companies that are the most secure tend to be the ones spending the least money on security. I get the feeling that shops which are closer to category 2 are going to read the Gartner summary and decide to cut their IT security budgets in half in hopes that fixes all their problems, instead of investigating the real underlying issues: hiring competent people who can do security.
I agree, I don't see any reason for all the GoogleBacklash we're seeing recently. Google is still a fine company that employs a whole lot of really smart guys that I greatly admire, and lets them run wild doing really cool things The Right Way. I have no problem with how they're operating GMail, as a matter of fact I love it. I'm not so sure about this browser thing - I'm hoping it's just skinning and Google-related extensions on Firefox, or at least Gecko-based in general.
Cisco is the de-facto brand of networking gear for standard stuff like Ethernet. How much better are these high-performance switches people are talking about as suggestions in the comments here? This is not a rhetorical question, I just realy want to know and I'm too lazy and uninterested to look into it myself, but not lazy enough to stop typing this slashdot post. Is it enough to be worth going with non-Cisco for HPC clusters that use Ethernet-based interconnects? I know Cisco isn't infallible, but for all kinds of reasons they're a good bet in networking gear come purchasing time, at least outside this HPC cluster business.
Err.. well, I was half blind - the crypto benchmark part of the article does read that the 64-bit binaries are faster, I think, which agrees with the graphs - I just misread it after reading the conclusion first. But in the conclusion, they make the statement "Unfortunately, there were still several cases where 64-bit binaries performed slower; John the Ripper and OpenSSL being two of those examples."
The conclusion is in error, I think..
Maybe I'm blind (seems to be the case), but I stared at the OpenSSL graph results, and I see the opposite of what the written text claims about them both on the benchmark page and in the conclusion. The written statements were to the effect that the 64-bit binaries sucked and that it was probably because OpenSSL was so heavily 32-bit optimized - but when you mouse-over for the 64-bit OpenSSL graphs of AES and RSA, the 64-bit binary result numbers look like they're trouncing the 32-bit counterpart binary on the same processor, as well as everything else in the test. What gives?
If they put out an ad right now offering $100k+ for any bigtime spammer (or with a spam sample saying "Get the guy who sends *these*), I would probably be among many people who would track these people from their spam all the way back to their home addresses and fly out there and physically confirm it and haul them in if we had to, for that kind of money.
More importantly, even if all of the above issues were addressed, you're really getting a picture of how internet-connected probably engineering minded individuals sprinkled around the world that happened to catch wind of this site would have voted, which is nothing like the true population of the entire world would vote, I'm sure.
Very elegant, google would be proud
I was under the impression that this earlier UT "long nanotube" was actually made by braiding smaller nanotubes together into a larger "strand". The difference here is that the 4cm world record is for a single tube, no braiding things together.
Well, I'm not a scientist of any type, much less a physicist. I guess my layman brain just happens to be more aligned with the physicists' worldview
After reading your post, I'm inclined to think that the whole point of science is to construct a reductionist set of explanations that branches from the lowest levels of physics upwards to explaining everything else that needs explaining. Science is just how we conduct experiments in an attempt to flesh out this reductionist knowledge.
What happens when you farm enough wind power to replace all the fossil fuel burning that goes on in the world today? Does it suck enough life out of big wind current patterns to affect them in any significant way? (Which could, in theory, cause storms, sea level changes, etc...)
His point is that you're a crackpot. I've read your writing too. You want to know why all the other sciences deal directly with causality and physics doesn't? I'll tell you. It's because all of the other sciences are essentially specialized subfields of physics, and generally abstracted to a higher level than physicists work at. Ultimately, in any other branch of science, when you follow the scientific chain of causality, it leads back to physics, every time. Physics is the study of the lowest-level functioning of the reality we live in, whatever it is. You can only ask for a causal Why at the lowest layers of current physics (at any given time) for two reasons - either as a philosophical side-track that has little to do with furthering physics, or to prompt the unpeeling of another layer in the physics onion (Why do protons and neutrons and electrons behave as they do? Aha, it's those damn quarks, we've unravelled another onion layer, yay for physics, and so forth). Eventually we'll reach the very bottom, if one exists.
Well, perhaps my personal views are different than yours on this matter. But just as you gave the example of being able to be good freinds with someone of opposing worldview, I believe it's possible, and perhaps refreshing an interesting, to do the same in a LTR. I'm not saying one should seek that out, but I am saying that I would let chemistry and whatever else decides who ends up with who be the primary factors, and political alignment would be pretty low priority. I suppose if the worldviews were at opposite ends of the spectrum and both individuals were sycophants of their political faith, allowing it to dictate all aspects of their life, then it clearly wouldn't work, but then I would think those people are giving undue importance to such matters, as I said.
You're right of course, but I bet if you did as a block-level jit thing and kept the original asm code around in memory too, you could work it all out. Sounds like something Transmeta probably had to work through in their translating microcode I'm sure.
It's not that simple, which is why other projects like Wine haven't taken that approach. Ultimately some software ends up touching things in a hardware-specific way that the kernel does not abstract for you (it merely provides a pass-through interfaces for accessing some hardware in these cases), in which case the library doing so has to be translated as well.
Sadly enough, legal matter are the most important ones. If you value your own freedom enough that you don't want to risk ending up in federal pound-me-in-the-ass prison, then you really have only two options: Either learn the basic of the legal system regarding self-defense and deadly force, or choose to go unarmed and defenseless so that you'll be less likely to accidentally cross those laws the wrong way. And of course, if you value your own freedom, then option two really isn't an option.
I think you're probably intelligent enough to realize that there's an extremely low probability you'll actually be elected President of the United States this year. Do you think the Libertarian Party will ever win a Presidential Election? If so, what's the plan for getting there, and how long do you think it will take?
Actually, I could conceive a brilliant software engineer coming up with a universal translation mechanism that turns x86 assembler into functionally equivalent PowerPC assembler, or vice-versa, or to other platforms. I believe IBM had been funding research in thsi area for quite some years now.
What sets off the BS detector for me is the APIs. They consistently state that they can do this for any OS. You have to do API translation, and you'd have to do that per OS, and it's a staggering volume of work to get all the APIs translated (think Wine project, just trying to do windows->linux api on a single shared hardware platform). When my linux binary calls any given kernel, C library, or even other common library (readline?, pthreads?, opengl?, etc..), those calls all have to be translated to equivalent MacOS or Windows API calls.
SBC is a baby bell in heredity. I'm assuming TDS Metrocom is a CLEC, and that they are in fact using re=sold SBC lines, since I haven't heard tell of any true alternative LECs running their own physical networks. Your lines are still serviced and providiosned by SBC, subcontracted to TDS Metrocom. TDS is probably just operating a more efficient business on a lower profit margin than the front offices of SBC do.
You don't want 50 seperate startup companies all laying their own custom fiber or coax networks through your city redundantly, when you know that when it all shakes out, at best 3 will survive. Because physical infrastructure is involved, typically communities award a contract to a single player, who will provide said infrastructure - and they tend to keep that contract going because of the infrastructure (there would be financial issues switching to a new provider and possibly having to pay for them to buy the old infrastructure from the previous company).
The government's answer to these competitive problems to date in the electric and telecomm markets is to enforce a really stupid form of competition, and require incumbent telcos and power providers to share their physical network with startups (where "share" means basically resell). So now You can choose Traditional Company A, or one of A's 50 crappy resellers who have no physical infrastructure of their own.
The Right Thing, IMHO, is that municipalities (or states, or whatever) should be putting out bids and taking contracts to build physical cabling infrastructure of various types for the area, and also contracts (perhaps from the same provider, but not neccesarily) to maintain said cabling infrastructure. The cabling infrastructure is then owned by the municipality itself, and it terminates at certain wire-centers where competitive providers can hook in with their own real equipment (much like internet NAPs in some ways - a shared facility where people co-loc edge equipment).
Anyone else find it odd that since this story it's been an abnormally long time for /. to go with no new stories on the front page? Did the editors all go out drinking together or something?
I don't know who's confused about what here, but I'll run down the relevant laws right quick to clear this up:
1) NFA (1934) - Esentially "banned" full-auto weapons for civilian use, as well as a whole lot of other things (heavy duty explosives, big tank rounds, grenades, etc... you get the picture, silencers, too, I believe are covered). The reason I put "banned" in quotes is because it didn't really ban this stuff - it just required that you get a proper license from the Federal BATF for each weapon on this list you wanted to own, and pay them a tax. They made getting that license very difficult for most people, which makes it effectively, mostly, like a ban. The key point in the process of getting the license is that you have to get a local head of law enforcement (police chief, county sheriff, etc) to write in a personal letter stating that they authorize your ownership of a full-auto weapon, and they know you and love you, and think your full auto weapon will be a great asset to the community (good luck with that one, especially if you're not an upstanding member of your community to begin with - the vast majority of police chiefs wouldn't write that letter for their own officers' personal gun collections, much less your ass). Note that the license is per weapon, you have to do the whole process over again for each individual gun (or silencer, or mortar, or whatever your banned item of choice is).
2) The 1986 law (whatever it's called) - this went a little further and said that civilian ownership of any full auto manufactured post-1986 was flat out illegal, and you can't even get a license to cover it. So now, pre-1986 full-auto are ok (if you have the license, which you're pretty unlikely to have), but post-1986 full-auto guns are always illegal, no exceptions.
3) The 1994 Clinton Gun Ban (which expires tonight at midnight, thank god) - was just about completely pointless and didn't change any of the above. It has nothing to do with full-auto weapons, only semi-auto weapons which are functionally identical to many non-banned "hunting" rifles. And it was cosmetic in nature. Statistics over the past 10 years have proved that it had zero practical effect on crime or collecters' gun stockpiles of varioues categories of weapons, and it's finally going off the books.
Now as a side note, it is useful to realize that there's a loophole of sorts in the original 1934 "ban". The whole thing about getting police chief signoff does not apply to you if you happen to be a corporation properly registered and operating legally within your state. Therefore it is conceivable and doable (many have done it) to set up a personal corporation, file for the license under the corporation's name, and get your license to own those restricted weapons. The 1986 thing still applies of course, so you have to buy ones manufactured (and registered as NFA firearms in the US, I should point out) before that date - which still means no Glock 18's. Since there is a limited pool of pre-1986 full-auto weapons that are properly NFA registered in the US, and a great many gun nuts and collectors, the prices of these pre-86 full-autos is sky high to boot. You can get a cheap crappy submachine gun that was converted from semi-auto (and registered converted as NFA before 86...) for a few thousand if you're lucky, but if you want a Brand Spankin New In Box Colt M16A2 (mfg before 86, but never fired or messed with) today, expect prices upwards of 12-15k.
Yeah I should've mentioned the higher percentage stuff. I usually buy it at the pharmacy in the same section as the regular "rubbing alcohol". The purer bottles are usually labelled for use with hypodermic needles (to swab the area before you stick the needle in).
The problem is mostly a legal one, as anyone who's taken a concealed handgun course will tell you. The laws vary by state, but many are like Texas, where I'm at and can reasoanbly describe.
One factor is threat escalation. When one party initiates the use force, then the other party is justified in using force as a defense. When the first party escalates to deadly force, then the other party is justified in using deadly force in defense. Whoever initiates each escalation is the agressor who will likely be criminally convicted, whoever defends without escalating further up the chain of "nothing->force->deadly force" than the other party has already done is in the clear on defensive grounds. Where this all ties back in to the point is that while drawing a weapon and pointing it at someone only constitutes "force" (and is therefore legally no different than shouting, pushing, or grabbing their arm), firing the weapon, even into the ground or air as a warning shot, constitutes deadly force.
In Texas in particular, property rights are strong, and you can initiate deadly force in response to certain property crimes under certain condititions justifiably, even though the person committing the property crimes isn't neccesarily using deadly force against you. It's complicated, but a good rule of thumb for this stuff is that if it's dark, the guy is either inside your house and not clearly visible (lights are out), or it's dark and the guy is in your lawn showing signs of attempted arson (gas can in hand), you can shoot.
But just as one should never fire a warning shot (as it is a meaningless threat escalation and puts you on the wrong side of certain legal issues), one should also never "shoot to kill", or at least never phrase it that way to the cops who show up afterwards or the grand jury you'll be facing even in a defensive case. The important thing is that you were "shooting to stop" (either stop the property crime under the right circumstances, shooting to stop the threat on your life or that of others, shooting to stop "deadly force" actions against you or others, or any combination of the above). That also means that once the person does "stop" (dead, severely disabled and immobilized, running away, whatever), you are obligated to stop shooting, or once again you're on the wrong side of the law. Therefore a practical consideration to keep in your head (But enver say out loud) is that when you do make that shot to "stop", it better be deadly - because chances are after the initial hit the guy *will* stop one way or the other, and if he stays alive, he's likely to sue in civil court for his enduring medical problems if you winged him.
So, in summary, make sure you know your state's laws about when the use deadly force is authorized (A concealed handgun course in states that offer it is a great source of information and training) - and shoot to kill, but don't ever admit to shooting to kill, only shooting to stop - and do stop shooting when they stop aggressing. I would personally recommend a double-tap to the chest for your opening and closing volley.