They said "I want in-car coffee maker". In order to provide Wifi access or any other stupid gizmos, you must have something that will make coffee in a car. Expand from there. If they come back later with "Haha, you could just plug a Mr. Coffee into a rectifier and put it in the passenger footwell, then give me Wifi access", then you kick the interviewer in the crotch.
There is no way to get around having systems as described by the interviewee, no matter what crap you later add on top.
Re:The article sounded reasonable until:
on
A Gamer's Manifesto
·
· Score: 1
Real AI code is branchy. In order execution doesn't allow for moving instructions around to fill branch delay slots. Therefore, performance on branchy code is reduced by pipeline stalls. Perhaps a tightly targeted compiler can compensate. But due to the dynamic nature of the code, I think that is unlikely.
Maybe there's speculative (predicated) execution, though I don't remember that PowerPCs have that.
Are those "special" defense companies that don't have to employ, you know, U.S. citizens who can get one those...whaddayacallits... security clearances?
No, I don't think you'll see defense companies looking for H1B applicants. Unless we stack them like cordwood and burn them for warmth, they aren't of much use: even a janitor has to be able to get a clearance.
When you are doing something that is (a) important and (b) final, you usually do something like a "Consent to Ship", where you check over all the details of things with neutral experts. One of the things you should have at these events is a hardware and software pedigree: what is installed where and what is the trail that led to this happening (software revision numbers, hardware serial numbers, and so on). If they'd had an audit trail like this, and then checked their serial numbers against the actual hardware, they'd have figured this out.
That they did not speaks to me of a rather slipshod process. I understand that space probes are not widgets that are churned out by the dozens, so maybe a perfectly polished list of tasks is not to be expected, but exactly *because* of the uniqueness of the mission, they should be careful.
So, are we sure this is the only screw-up? Or is the only *detected* screw-up?
Of the $55 billion in cuts, they then redirect $25 billion to the Army (mostly for Iraq-type support stuff, I think).
So it's $30 billion in cuts, which is still a decent amount. But I dunno how much of that will survive, since $18billion of that affects LockMart, and I have never ever seen political operators smoother than they. I'm not sure that Georgia's Congressional delegation will allow F-22 to be cut, and I'm sure some other (or the same) delegation will feel the same way about the C-130J, and the Virginia attack subs, and retiring the Kennedy carrier, etc.
Those things employ a *lot* of people, and no wants to have to deal with that. Dunno. I think the Kennedy will go despite any objections from Florida (where its based), but that leaves the 4 remaining carriers based on the east coast all at Norfolk, so they'll try to steal one of those, but Warner (Virginia senator) is head of the Senate Armed Services Committee (I'm pretty sure), so how that could happen... Plus the Norfolk carries are all nuclear, whereas the Kennedy is not, so there'd have to be a lot of infrastructure changes to handle a nuke in Florida.
If you're a "semi"-public agency, then some part of your budget should be in filling FOIA requests. Because, damnit, I'm paying you to generate the documents you then want to charge me to retrieve.
It would be completely evil to force someone into bankruptcy to find out about their brother's death in prison (the cause for the most recent FOIA controversy, I believe). And hell, with that excuse, I can easily price any information I don't want to release to you out of your price range by generating spurious documents or whathaveyou.
"Oh, you want to know what Ashcroft had for lunch a year ago? Well, those documents are on 48 billion sheets of paper (he's a hearty eater), and we'll need a billion dollars for our 40 copying specialists ($300 an hour) to run those pages through our gold-plated Xerox machine. Is that cash or credit?"
So moan all you want about what it takes to service requests from the public, but any agency should (a) be prepared for such requests (b) really be prepared for such requests.
This doesn't excuse absurdly broad requests, but at then you could reply "Narrow your search because we estimate 2 million manhours and $400k in expenses".
Uh, many many military folks use the word warfighter. It describes the jobs of the front-line combat service personnel from all branches of the military (not cooks or other non-combat personnel).
"We provide the best XYZ possible to our nation's warfighters".
I've seen it for a number of years in (a) magazines devoted to military equipment (Journal of Electronic Defense (JED)) (b) heard it from the various military customer-type people I come into contact with as a defense contractor (they may have been infected with 'bizspeak', though, for all I know).
JED in particular has a column called "I: First Person Singular" that is usually accounts of retired service personnel describing their experience with various electronics defense systems (radar, ECM, ECCM, other EW gear) in combat situations (World War II, Korea, and Vietnam). The people writing these columns often use warfighter without any indication of discomfort or irony.
It's a real word that is just now percolating into common usage via the enlightened interest in things military spurred on by the conflict in Iraq.
They can *tell* you that there're metal threads running through the cover, but can you know that without dismantling one? Perhaps the activation frequencies will be made public, but perhaps not. In any event, it would probably be a pain in the ass to figure it out non-destructively (try and stuff an antenna in there and keep the passport closed, then measure the intensity of the radiation that comes though? Microwave it and look for sparks or the wires to catch fire?).
Make my tinfoil hat a beanie with a propellor, please. Or maybe a fedora...
Ha ha, you only wish we upgraded them. Mostly you just wait for the last people who knew how they worked to retire or die, then you burn the lab to the ground and start over.
I kid, I kid.
Oh wait, no I don't.
ECL logic and analog computers are not that uncommon, though. Know Occam? Be welcome! Learn to read F77 written in all caps and only 70 columns wide (gotta fit the punchcards!).
What I work on is among the newer stuff, and why yes, I do use SGIs. But SGI is hard at work cannibalizing this market, so I imagine that there is maybe one more generation for their machines, and that's it. Why? Well, the reason that we use SGIs is because of the good realtime capabilities of Irix. And Irix only runs on MIPS, and MIPS has been EOLed (End-of-Lifed) in favor of Linux on IA-64. So either we see realtime Linux capabilities from SGI which will have to be folded back in the kernel (and then I'm not buying an SGI machine because I loathe IA-64). Or we see realtime Linux from other people (Finite State Machines Labs, or even from Ingo Molnar's regular kernel work), and I don't need SGI's work, and then I'm *still* not buying an SGI.
Either way, they're out. And I *want* them to be gone, because I hate their tools, hate their support, and hate their hardware. And I'm almost certainly their most knowledgeable and technically proficient customer in the desert SW. Guess what 8GB of RAM (PC3200, except willfully incompatible with PC DIMMs) cost: $25,000. That's right: basically was given a choice between a nice car and 8GB of commodity crap RAM.
They still do this. The current demo runs on three very large flat panels, and is a simulation of the Earth as seen from space. You can zoom in on any point on the globe, and as you zoom they add location-appropriate details from a massive disk array (I think they call it InfiniteStorage) at some tremendous rate (a couple Gigabytes a second, I think). It was pretty neat, but completely inappropriate for what I do.
I've been running Debian on the same SMP machine since 1998 (since 2.0.30, I think): 6 years without a reinstall. I think it was potato or hamm or something back then, but though I've swapped out the motherboard and processors a few times since then (never the disk the OS is on), and had a few things break (a motherboard that then took out 2 processors, and before that a disk), that machine is the one I still use today (dual 1GHz Pentium IIIs).
I know a lot of people who run Linux, but I know of no one who can say that they have gone as long without a re-install. Heck, I have a puny 166MHz Pentium running Debian stable as a gateway, and the only reason it hasn't spun past the 497 day uptime rollover (for older kernels) a couple times is power outages (it runs 2.2.x, so it's not like I've missed kernel updates).
My only fear for the future is that the old 4GB Seagate SCSI disk (full height!) will finally give way, or I could go a decade with only the occasional apt-get dist-upgrade (I run 'testing').
That's why Debian is cool. My machines are for doing work, not testing toys.
It's pretty wacky. The commonly used type of laser is Nd:YAG -- Neodium-doped Yttrium Aluminum Garnate (without the spelling errors). This is because the laser frequency is 1.064um, which is where the atmosphere is permeable. If you look at an absorption spectrum diagram, it's pretty hard to get stuff to go very far through the atmosphere with any intensity -- *especially* around the absorption frequencies of water.
Anyway, Nd:YAG is good at this, and is used in laser designators by the military for this very reason. I don't remember what the Airborne Laser is supposed to use for its three different lasers, but then they have a 747 to carry their stuff. So, to conclude -- it's hard to shine stuff through the atmosphere, and Nd:YAG are not in the visible spectrum (they're IR lasers), so no pilots are seeing these things even with smoke in the cockpit.
Oh, and from experience, the spot cast by a laser designator from a couple of hundred meters away on a T-72 is about a meter wide. It'd be huge at say 30,000 feet -- I don't know how it could hurt anyone at that range.
Patriot stores time in a 24-bit floating point number. Round-off and truncation errors accumulate over time.
It was a limitation of the hardware of the time, probably matched to cost constraints based on requirements from the Army as to the necessary uptime a Patriot battery should support.
You cannot fault Patriot here, since it functioned as designed and the Army was aware (even if the operators of the batteries were not informed) of the Patriot uptime limitations. In missile system design, a lot of decisions get made based on the requirements from the customer: picking a battery based on power consumption and required missile endurance, picking rocket motors based on required shelf lives (15 years or more!), etc.
Contrast this with Windows, which simply overflows a 32-bit counter of milliseconds (49.71 days). There really shouldn't be a limitation like this, even in a desktop system.
Not all of them. Vandenberg is the launch site for the targets used in testing EKV (uhm, mid-Phase Intercept or whatever they're calling it now). Targets out of Vandenberg, interceptors out of Kwaj.
A few figures will show this in a slightly different light. This is my business, so I have some idea what I'm talking about:
All of Europe put together (the whole EU), will spend about $200 billion (U.S. dollars) on defense next year. The U.S. budget looks to be about $420 billion next year. So the U.S., by itself, is spending more than twice as much as the entire continent of Europe.
A lot of the European spending is duplicative: i.e., everyone's got their own kind of tank, their own brand of artillery piece, often their own fighter (Eurofighter vs. JSF investment, or Viggen, or Dassault), etc.: this means that the the $200 billion is effectively a lot less than that.
Europe has no means to transport it's troops anywhere. They have no airlift capability at all. It would be faster to just have their guys pile into their own cars and motor to whereever they're going to fight and park. Europe has little ability to project power around the world: only now is France building a carrier or two, and Britain only has it's little jump carriers.
Galileo isn't up yet, and we'll have to see how well it works. Without it, Europe has no precision in projectile guidance without U.S. say so, except for IR or laser-guided munitions.
Most European arms are purchased from the United States. This could change, but everyone uses Tomahawk, everyone uses Sidewinder, everyone uses AMRAAM, everyone uses Harpoon (well, there's Exocet, but..), everyone uses TOW, everyone uses Javelin, everyone uses Patriot (thought MEADS is coming along: but that's also a mainly U.S. led effort). The list goes on.
And if you think railguns are the limit of what's coming, you'll be surprised. Coming are GPS-guided artillery shells, artillery shells with IR seekers in them, infantry-portable mortar rounds with seekers of one kind or another. And a huge push toward mechanization: robotic hunter-killer aircraft (Boeing) and tanks (the Army's Future Combat Systems projects). All these advances are thanks to the only country on the planet that really advances the state-of-the art in military technology: the United States. Not to be a dick (well, too much), but Europeans should consider this everytime they complain about the United States: the American public is footing the bill for almost the entirety of the world's military research.
Note that Maine spent $8,232 dollars per student on education in 2001. You think that'd be enough money to get a decent education, huh? Could it perhaps be being absorbed by a monstrous and corrupt bureaucracy that has enshrined the ideas of seniority and stability over competency? And whether you send your kids there or not, they still get your money: BONUS!
EIGHT THOUSAND FREAKING DOLLARS. My college tuition only started to be more than that my last year or so, and it's not like middle school students have the same facilities needs as a major public university. Craziness.
Note also another gigantic number. The United States spent $400 BILLION dollars on education. If the current population of the United States is 293,163,628 (and it is/was), then that's $1364 a year for every man, woman, and child in the country. That *should* be enough, don't you think?
Here's what I drive: a Nissan that I made sure was manufactured in Japan. Why? Because the quality of cars manufactured in the United States is quite atrocious in many cases. I have learned this from a coworker whose husband owns a used car dealership, and thus is always driving some new kind of vehicle (she considers anything with > 10,000 miles (16000 km) "high mileage"). He only sells American cars (some used and some bought new), and I cannot tell you how many Tahoes (>$40k per), Avalanches (similar), and various other pieces of American shit she has driven that develop incredible rattles and other mechanical annoyances and faults in the few months she has the cars. Brakes grumble, windows won't roll up and down, etc. -- and this coworker is typically *very* careful with the vehicles both because she is a tentative driver and because the vehicles must eventually be sold.
Of course it's not just American brands that are the problem, because obviously a lot of foreign marques make some of their models here (BMW, Mercedes, Toyota, etc.): I refuse to buy those, too. Look at the horrific quality with the Mercedes M-class, built in Alabama. Ugh. And these are built by those after the exodus: people who should be the cream of the crop and presumably thankful and industrious.
See, I'm not buying a car built somewhere else because it's cheaper: hell, I can only imagine the cost of labor in Japan is as high or higher, and besides Japanese, I'd only consider German makes built in Germany, and you *know* their labor costs are outrageous. But people in those countries seem to understand the importance of quality and taking pride in their work, something that often seems lacking in American cars. This is saddening, but to me it is the Truth.
What's worst of all is that, because my company was once owned by General Motors, I can receive a pretty damn substantial discount on a GM family car. So if I wanted a Saab, or one of the new GTOs, or potentially even one of the sweet new Cadiallc CTS-Vs, I could get one cheaply, but I won't.
I don't think the same kind of argument applies to the IT field: I think many of these people do very conscientous, quality work, it's just more expensive. And that's not really the argument being used for outsourcing (though it is used for H1-B visas, I suppose: finding technical talent abroad that "just isn't available here") anyway: cost, not quality is the argument being employed. If you want to make that argument, go for it, but I think it would be a difficult proposition to condemn an entire industry for sloth when the barriers to entry are so low and there is no organized labor trying to maintain a status quo (i.e., no overachievers allowed).
Perhaps for very old systems, this is the case, but no longer.
For instance, the U.S.'s Javelin missile (shoulder launched, IR-guided anti-armor) is what is known as a "wooden round": there is no software in the missile until it is attached to it's canister/launch unit (CLU). The CLU is the thing that the guy looks through in order to find and lock on to targets.
Anyway, when you attach the missile and it boots, it does so by reading all its software from the CLU. In this way, if you want to distribute an upgrade, what you need to upgrade is the image in the CLU, and not in every missile. Since there is probably only one CLU for every XYZ (6? 30? a goodly number) of missiles, this saves you a lot of effort.
This is the new standard way of doing things, at least in land combat type operations. I don't know much about how the air-to-air guys do it (AMRAAM, AIM-9x), but since you have to use a cart to move the missile and mount it on the plane before a flight, you could surely do it while in the depot.
I agree about the licensing and so forth. Mostly, though, the desire to use some form of Linux is that it can become customized in to a local dialect, or what have you. Like Lockheed Martin could have FighterLinux, which they've reviewed and customized, etc., and which fits their needs, while a major missile maker could have MissileLinux, which has been vetted. Local expertise is better than having to rely on vendor-supplied contractors and retards. And at least with Linux, if it came to that, you would have your choice of consultants.
I used to work in a public university's IT department, and the paper usage for many of the busyier (busier?) printers was a ream (500 sheets) in less than hour or so (much less, sometimes, to the limit of the 24 pages a minute or whatever they could do, plus the time to swap jobs: 25 minutes?): basically maxing the printers pretty much non-stop.
Therefore, we had some 4Si and 4L's that had close to *2 million* sheets printed out. There was some little menu you could scroll to and print out these statistics, some kind of status sheet that would tell you things like the printer's IP and how much RAM it had, etc. 2 MILLION sheets of paper.
I just measured a ream of OfficeMax paper @ 2.25 inches tall, which yields a stack of paper 1.7 miles tall! Plus they were Postscript printers.
5Si's were pretty good too, but almost certainly beyond the reach of a consumer.
Slashdot Mirror
Bah.
They said "I want in-car coffee maker". In order to provide Wifi access or any other stupid gizmos, you must have something that will make coffee in a car. Expand from there. If they come back later with "Haha, you could just plug a Mr. Coffee into a rectifier and put it in the passenger footwell, then give me Wifi access", then you kick the interviewer in the crotch.
There is no way to get around having systems as described by the interviewee, no matter what crap you later add on top.
Real AI code is branchy. In order execution doesn't allow for moving instructions around to fill branch delay slots. Therefore, performance on branchy code is reduced by pipeline stalls. Perhaps a tightly targeted compiler can compensate. But due to the dynamic nature of the code, I think that is unlikely.
Maybe there's speculative (predicated) execution, though I don't remember that PowerPCs have that.
Big computer, defense, and,
Are those "special" defense companies that don't have to employ, you know, U.S. citizens who can get one those...whaddayacallits... security clearances?
No, I don't think you'll see defense companies looking for H1B applicants. Unless we stack them like cordwood and burn them for warmth, they aren't of much use: even a janitor has to be able to get a clearance.
so no spoofing it unless you are really good at factoring large primes :)
Well, I dunno about you, but I can factor a "large prime" pretty damn quick. "Uh, what is 1 and the number itself, Alex."
Uh. int64_t && uint64_t?
I hump your leg.
And I anti-disagree.
When you are doing something that is (a) important and (b) final, you usually do something like a "Consent to Ship", where you check over all the details of things with neutral experts. One of the things you should have at these events is a hardware and software pedigree: what is installed where and what is the trail that led to this happening (software revision numbers, hardware serial numbers, and so on). If they'd had an audit trail like this, and then checked their serial numbers against the actual hardware, they'd have figured this out.
That they did not speaks to me of a rather slipshod process. I understand that space probes are not widgets that are churned out by the dozens, so maybe a perfectly polished list of tasks is not to be expected, but exactly *because* of the uniqueness of the mission, they should be careful.
So, are we sure this is the only screw-up? Or is the only *detected* screw-up?
That's not quite right.
Of the $55 billion in cuts, they then redirect $25 billion to the Army (mostly for Iraq-type support stuff, I think).
So it's $30 billion in cuts, which is still a decent amount. But I dunno how much of that will survive, since $18billion of that affects LockMart, and I have never ever seen political operators smoother than they. I'm not sure that Georgia's Congressional delegation will allow F-22 to be cut, and I'm sure some other (or the same) delegation will feel the same way about the C-130J, and the Virginia attack subs, and retiring the Kennedy carrier, etc.
Those things employ a *lot* of people, and no wants to have to deal with that. Dunno. I think the Kennedy will go despite any objections from Florida (where its based), but that leaves the 4 remaining carriers based on the east coast all at Norfolk, so they'll try to steal one of those, but Warner (Virginia senator) is head of the Senate Armed Services Committee (I'm pretty sure), so how that could happen... Plus the Norfolk carries are all nuclear, whereas the Kennedy is not, so there'd have to be a lot of infrastructure changes to handle a nuke in Florida.
Interesting times.
If you're a "semi"-public agency, then some part of your budget should be in filling FOIA requests. Because, damnit, I'm paying you to generate the documents you then want to charge me to retrieve.
It would be completely evil to force someone into bankruptcy to find out about their brother's death in prison (the cause for the most recent FOIA controversy, I believe). And hell, with that excuse, I can easily price any information I don't want to release to you out of your price range by generating spurious documents or whathaveyou.
"Oh, you want to know what Ashcroft had for lunch a year ago? Well, those documents are on 48 billion sheets of paper (he's a hearty eater), and we'll need a billion dollars for our 40 copying specialists ($300 an hour) to run those pages through our gold-plated Xerox machine. Is that cash or credit?"
So moan all you want about what it takes to service requests from the public, but any agency should (a) be prepared for such requests (b) really be prepared for such requests.
This doesn't excuse absurdly broad requests, but at then you could reply "Narrow your search because we estimate 2 million manhours and $400k in expenses".
Thanks a lot, damn it.
Lessee... 2005... getting older... on the worst day of the year!
Oh yeah, things are looking bright. Wait, is that a low-flying airplane I hea
Uh, many many military folks use the word warfighter. It describes the jobs of the front-line combat service personnel from all branches of the military (not cooks or other non-combat personnel).
"We provide the best XYZ possible to our nation's warfighters".
I've seen it for a number of years in (a) magazines devoted to military equipment (Journal of Electronic Defense (JED)) (b) heard it from the various military customer-type people I come into contact with as a defense contractor (they may have been infected with 'bizspeak', though, for all I know).
JED in particular has a column called "I: First Person Singular" that is usually accounts of retired service personnel describing their experience with various electronics defense systems (radar, ECM, ECCM, other EW gear) in combat situations (World War II, Korea, and Vietnam). The people writing these columns often use warfighter without any indication of discomfort or irony.
It's a real word that is just now percolating into common usage via the enlightened interest in things military spurred on by the conflict in Iraq.
They can *tell* you that there're metal threads running through the cover, but can you know that without dismantling one? Perhaps the activation frequencies will be made public, but perhaps not. In any event, it would probably be a pain in the ass to figure it out non-destructively (try and stuff an antenna in there and keep the passport closed, then measure the intensity of the radiation that comes though? Microwave it and look for sparks or the wires to catch fire?).
Make my tinfoil hat a beanie with a propellor, please. Or maybe a fedora...
Ha ha, you only wish we upgraded them. Mostly you just wait for the last people who knew how they worked to retire or die, then you burn the lab to the ground and start over.
I kid, I kid.
Oh wait, no I don't.
ECL logic and analog computers are not that uncommon, though. Know Occam? Be welcome! Learn to read F77 written in all caps and only 70 columns wide (gotta fit the punchcards!).
What I work on is among the newer stuff, and why yes, I do use SGIs. But SGI is hard at work cannibalizing this market, so I imagine that there is maybe one more generation for their machines, and that's it. Why? Well, the reason that we use SGIs is because of the good realtime capabilities of Irix. And Irix only runs on MIPS, and MIPS has been EOLed (End-of-Lifed) in favor of Linux on IA-64. So either we see realtime Linux capabilities from SGI which will have to be folded back in the kernel (and then I'm not buying an SGI machine because I loathe IA-64). Or we see realtime Linux from other people (Finite State Machines Labs, or even from Ingo Molnar's regular kernel work), and I don't need SGI's work, and then I'm *still* not buying an SGI.
Either way, they're out. And I *want* them to be gone, because I hate their tools, hate their support, and hate their hardware. And I'm almost certainly their most knowledgeable and technically proficient customer in the desert SW. Guess what 8GB of RAM (PC3200, except willfully incompatible with PC DIMMs) cost: $25,000. That's right: basically was given a choice between a nice car and 8GB of commodity crap RAM.
Die, damn you, die SGI!
They still do this. The current demo runs on three very large flat panels, and is a simulation of the Earth as seen from space. You can zoom in on any point on the globe, and as you zoom they add location-appropriate details from a massive disk array (I think they call it InfiniteStorage) at some tremendous rate (a couple Gigabytes a second, I think). It was pretty neat, but completely inappropriate for what I do.
I've been running Debian on the same SMP machine since 1998 (since 2.0.30, I think): 6 years without a reinstall. I think it was potato or hamm or something back then, but though I've swapped out the motherboard and processors a few times since then (never the disk the OS is on), and had a few things break (a motherboard that then took out 2 processors, and before that a disk), that machine is the one I still use today (dual 1GHz Pentium IIIs).
I know a lot of people who run Linux, but I know of no one who can say that they have gone as long without a re-install. Heck, I have a puny 166MHz Pentium running Debian stable as a gateway, and the only reason it hasn't spun past the 497 day uptime rollover (for older kernels) a couple times is power outages (it runs 2.2.x, so it's not like I've missed kernel updates).
My only fear for the future is that the old 4GB Seagate SCSI disk (full height!) will finally give way, or I could go a decade with only the occasional apt-get dist-upgrade (I run 'testing').
That's why Debian is cool. My machines are for doing work, not testing toys.
It's pretty wacky. The commonly used type of laser is Nd:YAG -- Neodium-doped Yttrium Aluminum Garnate (without the spelling errors). This is because the laser frequency is 1.064um, which is where the atmosphere is permeable. If you look at an absorption spectrum diagram, it's pretty hard to get stuff to go very far through the atmosphere with any intensity -- *especially* around the absorption frequencies of water.
Anyway, Nd:YAG is good at this, and is used in laser designators by the military for this very reason. I don't remember what the Airborne Laser is supposed to use for its three different lasers, but then they have a 747 to carry their stuff. So, to conclude -- it's hard to shine stuff through the atmosphere, and Nd:YAG are not in the visible spectrum (they're IR lasers), so no pilots are seeing these things even with smoke in the cockpit.
Oh, and from experience, the spot cast by a laser designator from a couple of hundred meters away on a T-72 is about a meter wide. It'd be huge at say 30,000 feet -- I don't know how it could hurt anyone at that range.
Patriot stores time in a 24-bit floating point number. Round-off and truncation errors accumulate over time.
It was a limitation of the hardware of the time, probably matched to cost constraints based on requirements from the Army as to the necessary uptime a Patriot battery should support.
You cannot fault Patriot here, since it functioned as designed and the Army was aware (even if the operators of the batteries were not informed) of the Patriot uptime limitations. In missile system design, a lot of decisions get made based on the requirements from the customer: picking a battery based on power consumption and required missile endurance, picking rocket motors based on required shelf lives (15 years or more!), etc.
Contrast this with Windows, which simply overflows a 32-bit counter of milliseconds (49.71 days). There really shouldn't be a limitation like this, even in a desktop system.
Trust me, I know about these things.
Not all of them. Vandenberg is the launch site for the targets used in testing EKV (uhm, mid-Phase Intercept or whatever they're calling it now). Targets out of Vandenberg, interceptors out of Kwaj.
It's red.
It gives the illegals out in the desert a fighting chance when the Hummers and so forth come upon them unawares.
It's kind of a new sport: advanced tactical U.S. immigration (Now with more Dodging!).
A few figures will show this in a slightly different light. This is my business, so I have some idea what I'm talking about:
All of Europe put together (the whole EU), will spend about $200 billion (U.S. dollars) on defense next year. The U.S. budget looks to be about $420 billion next year. So the U.S., by itself, is spending more than twice as much as the entire continent of Europe.
A lot of the European spending is duplicative: i.e., everyone's got their own kind of tank, their own brand of artillery piece, often their own fighter (Eurofighter vs. JSF investment, or Viggen, or Dassault), etc.: this means that the the $200 billion is effectively a lot less than that.
Europe has no means to transport it's troops anywhere. They have no airlift capability at all. It would be faster to just have their guys pile into their own cars and motor to whereever they're going to fight and park. Europe has little ability to project power around the world: only now is France building a carrier or two, and Britain only has it's little jump carriers.
Galileo isn't up yet, and we'll have to see how well it works. Without it, Europe has no precision in projectile guidance without U.S. say so, except for IR or laser-guided munitions.
Most European arms are purchased from the United States. This could change, but everyone uses Tomahawk, everyone uses Sidewinder, everyone uses AMRAAM, everyone uses Harpoon (well, there's Exocet, but..), everyone uses TOW, everyone uses Javelin, everyone uses Patriot (thought MEADS is coming along: but that's also a mainly U.S. led effort). The list goes on.
And if you think railguns are the limit of what's coming, you'll be surprised. Coming are GPS-guided artillery shells, artillery shells with IR seekers in them, infantry-portable mortar rounds with seekers of one kind or another. And a huge push toward mechanization: robotic hunter-killer aircraft (Boeing) and tanks (the Army's Future Combat Systems projects). All these advances are thanks to the only country on the planet that really advances the state-of-the art in military technology: the United States. Not to be a dick (well, too much), but Europeans should consider this everytime they complain about the United States: the American public is footing the bill for almost the entirety of the world's military research.
Take a look a this.
Note that Maine spent $8,232 dollars per student on education in 2001. You think that'd be enough money to get a decent education, huh? Could it perhaps be being absorbed by a monstrous and corrupt bureaucracy that has enshrined the ideas of seniority and stability over competency? And whether you send your kids there or not, they still get your money: BONUS!
EIGHT THOUSAND FREAKING DOLLARS. My college tuition only started to be more than that my last year or so, and it's not like middle school students have the same facilities needs as a major public university. Craziness.
Note also another gigantic number. The United States spent $400 BILLION dollars on education. If the current population of the United States is 293,163,628 (and it is/was), then that's $1364 a year for every man, woman, and child in the country. That *should* be enough, don't you think?
Here's what I drive: a Nissan that I made sure was manufactured in Japan. Why? Because the quality of cars manufactured in the United States is quite atrocious in many cases. I have learned this from a coworker whose husband owns a used car dealership, and thus is always driving some new kind of vehicle (she considers anything with > 10,000 miles (16000 km) "high mileage"). He only sells American cars (some used and some bought new), and I cannot tell you how many Tahoes (>$40k per), Avalanches (similar), and various other pieces of American shit she has driven that develop incredible rattles and other mechanical annoyances and faults in the few months she has the cars. Brakes grumble, windows won't roll up and down, etc. -- and this coworker is typically *very* careful with the vehicles both because she is a tentative driver and because the vehicles must eventually be sold.
Of course it's not just American brands that are the problem, because obviously a lot of foreign marques make some of their models here (BMW, Mercedes, Toyota, etc.): I refuse to buy those, too. Look at the horrific quality with the Mercedes M-class, built in Alabama. Ugh. And these are built by those after the exodus: people who should be the cream of the crop and presumably thankful and industrious.
See, I'm not buying a car built somewhere else because it's cheaper: hell, I can only imagine the cost of labor in Japan is as high or higher, and besides Japanese, I'd only consider German makes built in Germany, and you *know* their labor costs are outrageous. But people in those countries seem to understand the importance of quality and taking pride in their work, something that often seems lacking in American cars. This is saddening, but to me it is the Truth.
What's worst of all is that, because my company was once owned by General Motors, I can receive a pretty damn substantial discount on a GM family car. So if I wanted a Saab, or one of the new GTOs, or potentially even one of the sweet new Cadiallc CTS-Vs, I could get one cheaply, but I won't.
I don't think the same kind of argument applies to the IT field: I think many of these people do very conscientous, quality work, it's just more expensive. And that's not really the argument being used for outsourcing (though it is used for H1-B visas, I suppose: finding technical talent abroad that "just isn't available here") anyway: cost, not quality is the argument being employed. If you want to make that argument, go for it, but I think it would be a difficult proposition to condemn an entire industry for sloth when the barriers to entry are so low and there is no organized labor trying to maintain a status quo (i.e., no overachievers allowed).
Sufficient?
Perhaps for very old systems, this is the case, but no longer.
For instance, the U.S.'s Javelin missile (shoulder launched, IR-guided anti-armor) is what is known as a "wooden round": there is no software in the missile until it is attached to it's canister/launch unit (CLU). The CLU is the thing that the guy looks through in order to find and lock on to targets.
Anyway, when you attach the missile and it boots, it does so by reading all its software from the CLU. In this way, if you want to distribute an upgrade, what you need to upgrade is the image in the CLU, and not in every missile. Since there is probably only one CLU for every XYZ (6? 30? a goodly number) of missiles, this saves you a lot of effort.
This is the new standard way of doing things, at least in land combat type operations. I don't know much about how the air-to-air guys do it (AMRAAM, AIM-9x), but since you have to use a cart to move the missile and mount it on the plane before a flight, you could surely do it while in the depot.
I agree about the licensing and so forth. Mostly, though, the desire to use some form of Linux is that it can become customized in to a local dialect, or what have you. Like Lockheed Martin could have FighterLinux, which they've reviewed and customized, etc., and which fits their needs, while a major missile maker could have MissileLinux, which has been vetted. Local expertise is better than having to rely on vendor-supplied contractors and retards. And at least with Linux, if it came to that, you would have your choice of consultants.
Older HP printers are unkillable.
I used to work in a public university's IT department, and the paper usage for many of the busyier (busier?) printers was a ream (500 sheets) in less than hour or so (much less, sometimes, to the limit of the 24 pages a minute or whatever they could do, plus the time to swap jobs: 25 minutes?): basically maxing the printers pretty much non-stop.
Therefore, we had some 4Si and 4L's that had close to *2 million* sheets printed out. There was some little menu you could scroll to and print out these statistics, some kind of status sheet that would tell you things like the printer's IP and how much RAM it had, etc. 2 MILLION sheets of paper.
I just measured a ream of OfficeMax paper @ 2.25 inches tall, which yields a stack of paper 1.7 miles tall! Plus they were Postscript printers.
5Si's were pretty good too, but almost certainly beyond the reach of a consumer.
Old HPs == R0xor!