Which means...you run longer trains and more of them. And let's see, if each tank car has a 10^-whatever chance of derailing per mile traveled, what happens when there's more cars, more trains, and more wear on the tracks. Does the chance of derailment get bigger or smaller. This is not a trick question.
What? You mean you can't just say "satellites" and magically your data becomes the gold standard? You mean you actually need to calibrate your satellite data to ground stations? Impossible.
Next you're going to tell me that the buoys need to be taken out and calibrated to a NIST-traceable reference once or twice a year. That's crazy talk. Science=magic, and you must believe!
By all accounts things only started to go south when Bill and Dave left and were replaced by a series of bean counters with no sense for what gives a tech company positive buzz and positive sales growth. Carly being example 1. Meg being example 2.
What, you say transients matter more than first-order models in nonlinear dynamical systems? Impossible. That's not what they teach in undergraduate statistics, therefore it must be heresy.
I am. I have a job as an engineer in the military industrial complex. I've also been told to drop what I'm doing because of $BULLSHIT_ADMINISTRATIVE_REASON only to have to pick it up again a year or more later and waste time getting myself and the right people back on track. I've also seen my colleagues do the same, and I've seen all of get screwed by the fact that after $WAITING_PERIOD, the resources we had marshalled the first time around aren't quite so easy to marshal the second time around, especially when you pull the rug out from under people enough times, they don't want to work for/with you the next time when for real, I swear, we have the funding to finish it, promise. If it's true for the 10M programs I've worked on, it's true times a hundred for a billion-dollar power plant.
In my workplace of several thousand people spread over a complex of 10+ sprawling buildings, calls to 911 get piped directly to security so as to prevent the chance of an ambulance showing up without the guards knowing to let them through the gate.
of having to bend over backwards around regs that were written with the best of intentions and end up doing at least as much harm as good. In this case, the good is making it trickier to piss away money on bad software; the bad is disincentivizing looking for a good $10k solution in favor of piecing together three or so half-assed $3499.99 solutions.
I don't have the magic answer for how to prevent government waste, but having fixed caps like that across the board doesn't seem to be doing much good, because the waste is in making wrong priorities and going through the process to spend on them, not in overspending on the right priorities.
Lean code is always an issue. If your code incurs a x2 to x10 overhead associated with the virtual machine, that's either 2-10x the hardware you need to spend money on to achieve the same throughput as before, and 2-10x the electric bill for compute-intensive applications. If you're nowhere near the limit of your box, you don't notice. If you've got rooms upon rooms of computers doing the same thing, and you're writing your code in not C/C++, then you're wasting money.
Well, let's see. % means its a conversion code, l means the converted quantity is a long, i means its an integer, so a long integer, but e means it's a float to be converted to exponential notation. But it was supposed to be an integer. Does not compute.
Uh huh. And I'm sure it was all smiles and sunshine pre 2003. Not saying we didn't blow the execution to an extent that some people shouldn't be able to look themselves in the mirror, but some places just invite chaos by their very nature. Case in point: Egypt, Syria, Libya: not a single American boot on the ground before the "unpleasantness". One could make the case the same thing would have happened in Iraq regardless.
In either case. If they don't have their shit together for some modicum of success in one formal and/or quantitative discipline, they don't make good programmers. You do indeed need to be right every single time in rapid succession when your computer program handles millions of dollars of other people's money, doles out their medications, and keeps their credit card info secure, all to the nth degree if the program touches high voltage, high current, the throttle/brake of a car or the control surfaces of an airplane.
Name two and their consequences. And by consequences I don't mean one time annoyances like "it's cumbersome to write the init scripts" but actual things like "this language forces me to use double the memory or twice the cpu" and explain how systemd fixes it without introducing a worse one.
I've had a job now for about 10 years where a large fraction of the time I wear a software engineer's hat. Looking back now, I can point to a lot of design decisions in the software I work on that made me go "WTF?" when I first saw them as a young'un, but after having to contend with them for a good number of years, and thinking about how I would do them differently, I've come to the conclusion that the original WTF may be ugly and could use some polish, but the decisionmaking that produced it was fundamentally sound.
The more I hear about LP and systemd, the more it screams out that this guy just hasn't worked with Unix and Linux long enough to understand what it's used for and why it's built the way it is. His pronouncements just sound to me like an echo of my younger, stupider, self (and I just turned 30), and I can't take any of his output seriously. I really hope a critical mass of people are of the same mind with me and this guy can be made to redirect his energies somewhere where it doesn't fuck it up for the rest of us.
Yeah, but that's all latency, none of it is throughput. Maybe I buy your argument if you're talking about changing a few bytes in a file of size 4K or so, but if your file is megabytes or gigabytes in size (like a bigass complex-valued double precision matrix), then I don't think you necessarily want to shuffle all of it across the wire and back.
Clearly you need to check your privilege. Not everyone has the had the benefit of learning to reason like a human being and you need to be inclusive of the those who haven't by refraining from pointing out the contradictions in their worldview.
No, inflation is not good. Deflation is not good. A government/central bankthat spends its time actually understanding the how money flows through the economy (as opposed to spending its time convincing everyone that it understands these things) would aim for a combination of interest rates, QE, and out-and-out printing money that makes the inflation as near zero as possible with maybe half an epsilon's worth of safety margin in the positive direction to keep the banks from freezing up.
I was responsible for the servo of an optical tracking mount with moving dome and powered cable wrap (no manipulator arms, just four axes of motion, three of them coaxial) and I still made sure to pump out about 100kbytes/sec worth of telemetry for all the moving machinery that was there. A 5 or six axis robot should probably be pumping out at least that much of telemetry.
The second real question is what their data retention polcy is so that human error can be isolated from electromechanical fault and software fault.
Even the multi-function displays in the middle of the instrument panels on *all* cars made in the last three or four years is too much. Old fogeys like myself, at the crusty old age of 29, have gotten used to associating a particular spatial location in an automobile's console with a particular piece of information so that it's second nature.
This is how the mind is wired to absorb information from the world at a very basic level. Want to see what the weather it is? Look up. Want to see if you're walking on steady ground? Look down. Want to see if there's danger or prey out there? Look around.
Same in a car, or fighter jet for that matter: Want to see the time? Look at where the clock is. Want to see what radio station you're listening to? Look at where the tuner is. Want to see how much gas you've got? Look at where the fuel gauge is. This is constant-time lookup. If you have multifunction displays that *change* where these basic things are, now you've upped the cognitive load on the driver in that he now has to keep track of what state the display is in rather than just glancing in a well-remembered spot.
A proper heads-up display, and a proper desktop GUI, smartphone app, etc, preserve this feature so that you can see what you need by looking where you remember. Incidentally, this is a large part of what 'type rating' is on commercial aircraft, and aircraft manufacturers frequently retain large commercial customers by laying out the cockpits of their newer models the same exact way as the old one, with the selling point that pilots don't need to be retrained to figure out where they need to be looking and where their hands need to be in the new cockpit.
The point is, a good HUD for a car will show the same thing in the same place all the time. Just projections of dials and needles if I had my way. No popups, no text to read, no nothing. If there's something wrong with the car, a single idiot light that says 'check engine' will do it, because you're not going to diagnose it yourself while on the highway. That way it actually does save you time and keeps your gaze closer to the road.
But yeah. If you've got bells and whistles and distractions in your field of vision, of course it's unsafe. Most people are probably smart enough to ignore the popup message crap polluting automotive mutlifunction displays, by keeping their eyes up. If the crap follows them there, that's not an usafe display mechanism, that's unsafe human interface design. </rant>
Slashdot Mirror
Which means...you run longer trains and more of them. And let's see, if each tank car has a 10^-whatever chance of derailing per mile traveled, what happens when there's more cars, more trains, and more wear on the tracks. Does the chance of derailment get bigger or smaller. This is not a trick question.
Still does. Light is a continuous electromagnetic (or probability amplitude) wave travelling through a glass waveguide.
What? You mean you can't just say "satellites" and magically your data becomes the gold standard? You mean you actually need to calibrate your satellite data to ground stations? Impossible.
Next you're going to tell me that the buoys need to be taken out and calibrated to a NIST-traceable reference once or twice a year. That's crazy talk. Science=magic, and you must believe!
By all accounts things only started to go south when Bill and Dave left and were replaced by a series of bean counters with no sense for what gives a tech company positive buzz and positive sales growth. Carly being example 1. Meg being example 2.
What, you say transients matter more than first-order models in nonlinear dynamical systems? Impossible. That's not what they teach in undergraduate statistics, therefore it must be heresy.
I am. I have a job as an engineer in the military industrial complex. I've also been told to drop what I'm doing because of $BULLSHIT_ADMINISTRATIVE_REASON only to have to pick it up again a year or more later and waste time getting myself and the right people back on track. I've also seen my colleagues do the same, and I've seen all of get screwed by the fact that after $WAITING_PERIOD, the resources we had marshalled the first time around aren't quite so easy to marshal the second time around, especially when you pull the rug out from under people enough times, they don't want to work for/with you the next time when for real, I swear, we have the funding to finish it, promise. If it's true for the 10M programs I've worked on, it's true times a hundred for a billion-dollar power plant.
In my workplace of several thousand people spread over a complex of 10+ sprawling buildings, calls to 911 get piped directly to security so as to prevent the chance of an ambulance showing up without the guards knowing to let them through the gate.
of having to bend over backwards around regs that were written with the best of intentions and end up doing at least as much harm as good. In this case, the good is making it trickier to piss away money on bad software; the bad is disincentivizing looking for a good $10k solution in favor of piecing together three or so half-assed $3499.99 solutions.
I don't have the magic answer for how to prevent government waste, but having fixed caps like that across the board doesn't seem to be doing much good, because the waste is in making wrong priorities and going through the process to spend on them, not in overspending on the right priorities.
Shows what they know. It's common knowledge that nothin' 'll ever beat out the full-throated roar of an eight-cylinder gasoline engine.
Don't forget to mention where most of Linux source code comes from too, BTW.
Lean code is always an issue. If your code incurs a x2 to x10 overhead associated with the virtual machine, that's either 2-10x the hardware you need to spend money on to achieve the same throughput as before, and 2-10x the electric bill for compute-intensive applications. If you're nowhere near the limit of your box, you don't notice. If you've got rooms upon rooms of computers doing the same thing, and you're writing your code in not C/C++, then you're wasting money.
Well, let's see. % means its a conversion code, l means the converted quantity is a long, i means its an integer, so a long integer, but e means it's a float to be converted to exponential notation. But it was supposed to be an integer. Does not compute.
Uh huh. And I'm sure it was all smiles and sunshine pre 2003. Not saying we didn't blow the execution to an extent that some people shouldn't be able to look themselves in the mirror, but some places just invite chaos by their very nature. Case in point: Egypt, Syria, Libya: not a single American boot on the ground before the "unpleasantness". One could make the case the same thing would have happened in Iraq regardless.
Fast. Government procurement. Hahahahahahaha. Speaking as a deeply embedded cog in the military industrial complex.
In either case. If they don't have their shit together for some modicum of success in one formal and/or quantitative discipline, they don't make good programmers. You do indeed need to be right every single time in rapid succession when your computer program handles millions of dollars of other people's money, doles out their medications, and keeps their credit card info secure, all to the nth degree if the program touches high voltage, high current, the throttle/brake of a car or the control surfaces of an airplane.
Name two and their consequences. And by consequences I don't mean one time annoyances like "it's cumbersome to write the init scripts" but actual things like "this language forces me to use double the memory or twice the cpu" and explain how systemd fixes it without introducing a worse one.
I've had a job now for about 10 years where a large fraction of the time I wear a software engineer's hat. Looking back now, I can point to a lot of design decisions in the software I work on that made me go "WTF?" when I first saw them as a young'un, but after having to contend with them for a good number of years, and thinking about how I would do them differently, I've come to the conclusion that the original WTF may be ugly and could use some polish, but the decisionmaking that produced it was fundamentally sound.
The more I hear about LP and systemd, the more it screams out that this guy just hasn't worked with Unix and Linux long enough to understand what it's used for and why it's built the way it is. His pronouncements just sound to me like an echo of my younger, stupider, self (and I just turned 30), and I can't take any of his output seriously. I really hope a critical mass of people are of the same mind with me and this guy can be made to redirect his energies somewhere where it doesn't fuck it up for the rest of us.
Yeah, but that's all latency, none of it is throughput. Maybe I buy your argument if you're talking about changing a few bytes in a file of size 4K or so, but if your file is megabytes or gigabytes in size (like a bigass complex-valued double precision matrix), then I don't think you necessarily want to shuffle all of it across the wire and back.
Ignoring people's basic sense of fairness (he's got half my skills, but earns just as much) is something that tends to make capitalism difficult.
s/the Germans/one German on the record as having an interest in discovering exotic physics/
Clearly you need to check your privilege. Not everyone has the had the benefit of learning to reason like a human being and you need to be inclusive of the those who haven't by refraining from pointing out the contradictions in their worldview.
No, inflation is not good. Deflation is not good. A government/central bankthat spends its time actually understanding the how money flows through the economy (as opposed to spending its time convincing everyone that it understands these things) would aim for a combination of interest rates, QE, and out-and-out printing money that makes the inflation as near zero as possible with maybe half an epsilon's worth of safety margin in the positive direction to keep the banks from freezing up.
I was responsible for the servo of an optical tracking mount with moving dome and powered cable wrap (no manipulator arms, just four axes of motion, three of them coaxial) and I still made sure to pump out about 100kbytes/sec worth of telemetry for all the moving machinery that was there. A 5 or six axis robot should probably be pumping out at least that much of telemetry.
The second real question is what their data retention polcy is so that human error can be isolated from electromechanical fault and software fault.
That's what they said about the F-4 in the 60's. Then they had to put a cannon on it real fast after theory met reality in the skies over Vietnam.
Mod parent up.
Even the multi-function displays in the middle of the instrument panels on *all* cars made in the last three or four years is too much. Old fogeys like myself, at the crusty old age of 29, have gotten used to associating a particular spatial location in an automobile's console with a particular piece of information so that it's second nature.
This is how the mind is wired to absorb information from the world at a very basic level. Want to see what the weather it is? Look up. Want to see if you're walking on steady ground? Look down. Want to see if there's danger or prey out there? Look around.
Same in a car, or fighter jet for that matter: Want to see the time? Look at where the clock is. Want to see what radio station you're listening to? Look at where the tuner is. Want to see how much gas you've got? Look at where the fuel gauge is. This is constant-time lookup. If you have multifunction displays that *change* where these basic things are, now you've upped the cognitive load on the driver in that he now has to keep track of what state the display is in rather than just glancing in a well-remembered spot.
A proper heads-up display, and a proper desktop GUI, smartphone app, etc, preserve this feature so that you can see what you need by looking where you remember. Incidentally, this is a large part of what 'type rating' is on commercial aircraft, and aircraft manufacturers frequently retain large commercial customers by laying out the cockpits of their newer models the same exact way as the old one, with the selling point that pilots don't need to be retrained to figure out where they need to be looking and where their hands need to be in the new cockpit.
The point is, a good HUD for a car will show the same thing in the same place all the time. Just projections of dials and needles if I had my way. No popups, no text to read, no nothing. If there's something wrong with the car, a single idiot light that says 'check engine' will do it, because you're not going to diagnose it yourself while on the highway. That way it actually does save you time and keeps your gaze closer to the road.
But yeah. If you've got bells and whistles and distractions in your field of vision, of course it's unsafe. Most people are probably smart enough to ignore the popup message crap polluting automotive mutlifunction displays, by keeping their eyes up. If the crap follows them there, that's not an usafe display mechanism, that's unsafe human interface design. </rant>