Locks that are considered up-to-snuff, security-wise, have a tamper-proof CPU with integrated hardware crypto and a coil driver on a ceramic substrate, buried right next to the coil. The coil is wound in a couple of sections to prevent simple "poke and energize" schemes. The magnet wire is bonded directly to the circuitry on the substrate. The power to the locking mechanism is always there. All that's required to unlock the lock is a proper command, sent over a properly encrypted link. It's done for the same reason that modern credit card strip readers encrypt right next to the head assembly, and the chip readers encrypt right in the chip on the card. You really don't want your plaintext, whether it's the CC number or the on/off boolean to be any more exposed than necessary.
OK, so they are super-efficient, so they should be able to pay their workers a living wage, right? Right? They can shove their efficiency in the Waltons' butts, as far as I care. It's an enterprise set up for the benefit of the sociopathic owners. It's a public company in name only.
They may be competitive in the cloud service provider market, but they are such an unethical company that I personally simply refuse to deal with them.
Oh, I'm not talking about Mims's stuff - his stuff does work, I went through all of his books as a kid and built every circuit there. I don't recall any mistakes in the schematics.
I'm talking about the previous edition of Art of Electronics. It's all peachy until you go through it with a fine-toothed comb and build every circuit, and see what happens if you follow their advice. They are sometimes sloppy with their language, too.
So, you're saying that no one should ever have a drink with a meal when dining out?
I don't know what he's saying, but I'm saying that you're correct. You shouldn't have that drink and then drive. What's so fucking hard to understand about that?
I can't see any contractor worth his salt saying "I'll build that building based on a computer file that can be updated by remote push down"
Nice straw man, and this is "5 Insightful"? The fuck?
Who the heck is talking of remote push down or undocumented changes? Do you even live in the same world I do? When the plans are final and approved, every contractor - at least in the U.S. - already gets their paper copies based on digital files that were a part of the bid package. They can also get those digitally, and have their site workers use whatever digital tech they care for to view those if they prefer that over dead trees.
The rest of your post makes sense, but please stop with the imaginary problems. Nobody is optically copying blueprints anymore.
I know that we've all heard it before somewhere, but I'm firm in my belief that people really don't seem to have a clue how to format MS Office documents properly. You should not depend on spacing. You should style, anchor and otherwise set everything up so that when spacing does inevitably change, things still work. That's very much doable - heck, I've had wonderful, LaTeX-like lab reports done in ~1998 that gasp open fine and look right on LibreOffice 5 on Mac. We're talking of 17 year old.doc files with nice flow, equations and drawings, last edited on Windows NT 4 and Office 95...
As someone who sometimes has to use the POSIX "API", let me tell you this: it's a piece of a fucking joke. As much as it pains me to say it, the functionality provided by the core winapi, even with its idiosyncracies, is way ahead of anything POSIX lets you do. Sure, you do want to use a decent C++ wrapper around either winapi or POSIX, but still: POISX doesn't do the very basics of what you need to develop actually useful applications: you need to provide lots of your own, or library, code to get where you want to be, foundation-wise. People hail POSIX as something wonderful, but all I see is a bunch of very much anti-UNIX philosophy codified for no good reason. Even Linux APIs are much nicer than POSIX - the modern "everything is a file descriptor" is much closer to what Windows provides with its universal object handle metaphor. All the while POSIX has special approaches to everything: when you wait on one thing, you can't wait on another thing, everything needs special treatment - it's disgusting, frankly said.
POSIX api was designed by people who never seemed to realize what the needs of modern, responsive application and server design are.
Windows 95 would not run _any_ Windows 2 programs.
Not that it's very important, but are you sure about it? It did have support for 16 bit winapi applications... I did run plenty of Windows 3.x code on Windows 95...
Frankly said, I'd use Windows on RPi simply to have access to decent documentation. Linux is a very loosely tied bunch of stuff, there's no overarching design and no single source of documentation - heck there is really no documentation for a lot of things, apart from the code itself. As much as I enjoy reading code, sometimes I'd rather read English and be sure that the API won't magically change overnight...
I can't develop an iphone app without enduring a multi gigabyte download/install of "XCode"?? are you for real?
LOL. When you install any development platform, do you think that the compilers, libraries and "documentation" just materializes itself from thin air? You had to download all of that stuff anyway.
I think that this is bullshit. VS 2015 works fine on Windows 7, and I don't see what's so special about building for the arm targets that would require Windows 10. I build for ARM using VS 2105 on Win 7 already, I just haven't deployed the stuff on RPi yet... shouldn't be hard, but I really don't see what's the outrage here. People, let's try it before you spew nonsense, mmkay?
There was a guy somewhere higher up who could orchestrate the specs so that nobody below them knew what was going on. In complex systems it's fairly easy to engineer emergent behavior even if no subpart of the system shows that behavior alone. Source: wrote plenty of "easter eggs" in large systems that were there to detect hardware/firmware copying. No single part of the system would show the easter egg behavior, but when the pieces were all together in a particular fashion, you'd get the revealing behavior that wouldn't occur if the system under test wasn't a copy the protected design.
Dafuck?! EPA does not mandate stupid engineering. The automaker fucked up. What, their ansys or nastran licenses ran out and they couldn't fucking model what was going on, and/or experimentally validate the engine block/head design? This is a noob mistake - whether precipitated by stupid management, or noob designers, or both, who knows. Blaming it on EPA is going full retard.
Maybe, instead of worrying about it, we just accept the reality: small-engine diesel is nonsense. It makes no sense either environmentally nor economically. Just as everyone who should know it knows that small turbines are inefficient and don't belong on passenger cars (but make all the sense on large airplanes), we all know that small diesels don't belong on passenger cars. How much evidence would you need to understand that it's not about the regulation, but about the reality of these engines. They make zero sense when they're small. Sure they are reliable, but almost any car would be more reliable without majority of common emission controls.
The especially egregious aspect of that is: this fundamentally flawed testing regimen isn't free in terms of resources either. Not only it costs us, the car owners, money, but that money, and everything it's spent on downstream, is pure waste. The energy used to perform the tests? You might as well run the same number of kWh through a heater into the sea. The personnel costs, and the amortized resources they need? You could just dump those into the sea with the same overall benefit to the society.
The make-work tests are worse to the environment than none. All the resources and productivity they consume could be better spent elsewhere.
Given the market performance of automakers, the real end-to-end gross margins on cars are hovering in the vicinity of 0%. It's a good day when they are slightly above zero, instead of slightly below.
Slashdot Mirror
Be careful, because a single counterexample will prove you wrong. And eventually it will happen.
All of them? Get off your high chair.
We could, if we had a couple trillion trillion trillion trillion universes at our disposal.
Locks that are considered up-to-snuff, security-wise, have a tamper-proof CPU with integrated hardware crypto and a coil driver on a ceramic substrate, buried right next to the coil. The coil is wound in a couple of sections to prevent simple "poke and energize" schemes. The magnet wire is bonded directly to the circuitry on the substrate. The power to the locking mechanism is always there. All that's required to unlock the lock is a proper command, sent over a properly encrypted link. It's done for the same reason that modern credit card strip readers encrypt right next to the head assembly, and the chip readers encrypt right in the chip on the card. You really don't want your plaintext, whether it's the CC number or the on/off boolean to be any more exposed than necessary.
OK, so they are super-efficient, so they should be able to pay their workers a living wage, right? Right? They can shove their efficiency in the Waltons' butts, as far as I care. It's an enterprise set up for the benefit of the sociopathic owners. It's a public company in name only.
They may be competitive in the cloud service provider market, but they are such an unethical company that I personally simply refuse to deal with them.
Oh, I'm not talking about Mims's stuff - his stuff does work, I went through all of his books as a kid and built every circuit there. I don't recall any mistakes in the schematics.
I'm talking about the previous edition of Art of Electronics. It's all peachy until you go through it with a fine-toothed comb and build every circuit, and see what happens if you follow their advice. They are sometimes sloppy with their language, too.
So, you're saying that no one should ever have a drink with a meal when dining out?
I don't know what he's saying, but I'm saying that you're correct. You shouldn't have that drink and then drive. What's so fucking hard to understand about that?
I agree. Some of the circuits don't even quite work as advertised...
I can't see any contractor worth his salt saying "I'll build that building based on a computer file that can be updated by remote push down"
Nice straw man, and this is "5 Insightful"? The fuck?
Who the heck is talking of remote push down or undocumented changes? Do you even live in the same world I do? When the plans are final and approved, every contractor - at least in the U.S. - already gets their paper copies based on digital files that were a part of the bid package. They can also get those digitally, and have their site workers use whatever digital tech they care for to view those if they prefer that over dead trees.
The rest of your post makes sense, but please stop with the imaginary problems. Nobody is optically copying blueprints anymore.
I know that we've all heard it before somewhere, but I'm firm in my belief that people really don't seem to have a clue how to format MS Office documents properly. You should not depend on spacing. You should style, anchor and otherwise set everything up so that when spacing does inevitably change, things still work. That's very much doable - heck, I've had wonderful, LaTeX-like lab reports done in ~1998 that gasp open fine and look right on LibreOffice 5 on Mac. We're talking of 17 year old .doc files with nice flow, equations and drawings, last edited on Windows NT 4 and Office 95...
As someone who sometimes has to use the POSIX "API", let me tell you this: it's a piece of a fucking joke. As much as it pains me to say it, the functionality provided by the core winapi, even with its idiosyncracies, is way ahead of anything POSIX lets you do. Sure, you do want to use a decent C++ wrapper around either winapi or POSIX, but still: POISX doesn't do the very basics of what you need to develop actually useful applications: you need to provide lots of your own, or library, code to get where you want to be, foundation-wise. People hail POSIX as something wonderful, but all I see is a bunch of very much anti-UNIX philosophy codified for no good reason. Even Linux APIs are much nicer than POSIX - the modern "everything is a file descriptor" is much closer to what Windows provides with its universal object handle metaphor. All the while POSIX has special approaches to everything: when you wait on one thing, you can't wait on another thing, everything needs special treatment - it's disgusting, frankly said.
POSIX api was designed by people who never seemed to realize what the needs of modern, responsive application and server design are.
Windows 95 would not run _any_ Windows 2 programs.
Not that it's very important, but are you sure about it? It did have support for 16 bit winapi applications... I did run plenty of Windows 3.x code on Windows 95...
Frankly said, I'd use Windows on RPi simply to have access to decent documentation. Linux is a very loosely tied bunch of stuff, there's no overarching design and no single source of documentation - heck there is really no documentation for a lot of things, apart from the code itself. As much as I enjoy reading code, sometimes I'd rather read English and be sure that the API won't magically change overnight...
I can't develop an iphone app without enduring a multi gigabyte download/install of "XCode"?? are you for real?
LOL. When you install any development platform, do you think that the compilers, libraries and "documentation" just materializes itself from thin air? You had to download all of that stuff anyway.
I think that this is bullshit. VS 2015 works fine on Windows 7, and I don't see what's so special about building for the arm targets that would require Windows 10. I build for ARM using VS 2105 on Win 7 already, I just haven't deployed the stuff on RPi yet... shouldn't be hard, but I really don't see what's the outrage here. People, let's try it before you spew nonsense, mmkay?
Sure, but these diesels pollute way more than they should. EPA standards are IMHO not draconian at all. I, for one, like my air clean.
I very much doubt that you have all that many professional (licensed) engineers working on cars. It's simply not required.
There was a guy somewhere higher up who could orchestrate the specs so that nobody below them knew what was going on. In complex systems it's fairly easy to engineer emergent behavior even if no subpart of the system shows that behavior alone. Source: wrote plenty of "easter eggs" in large systems that were there to detect hardware/firmware copying. No single part of the system would show the easter egg behavior, but when the pieces were all together in a particular fashion, you'd get the revealing behavior that wouldn't occur if the system under test wasn't a copy the protected design.
This. This a 100 times.
Dafuck?! EPA does not mandate stupid engineering. The automaker fucked up. What, their ansys or nastran licenses ran out and they couldn't fucking model what was going on, and/or experimentally validate the engine block/head design? This is a noob mistake - whether precipitated by stupid management, or noob designers, or both, who knows. Blaming it on EPA is going full retard.
Maybe, instead of worrying about it, we just accept the reality: small-engine diesel is nonsense. It makes no sense either environmentally nor economically. Just as everyone who should know it knows that small turbines are inefficient and don't belong on passenger cars (but make all the sense on large airplanes), we all know that small diesels don't belong on passenger cars. How much evidence would you need to understand that it's not about the regulation, but about the reality of these engines. They make zero sense when they're small. Sure they are reliable, but almost any car would be more reliable without majority of common emission controls.
The especially egregious aspect of that is: this fundamentally flawed testing regimen isn't free in terms of resources either. Not only it costs us, the car owners, money, but that money, and everything it's spent on downstream, is pure waste. The energy used to perform the tests? You might as well run the same number of kWh through a heater into the sea. The personnel costs, and the amortized resources they need? You could just dump those into the sea with the same overall benefit to the society.
The make-work tests are worse to the environment than none. All the resources and productivity they consume could be better spent elsewhere.
Given the market performance of automakers, the real end-to-end gross margins on cars are hovering in the vicinity of 0%. It's a good day when they are slightly above zero, instead of slightly below.
Why are you making shit up?