Very cool resource. I didn't know that lossy codecs relied on an "uncertainty principle" like relation, but it seems likely that hearing is not linear. I don't think that anyone seriously claimed that it was *impossible* to tell the difference between lossless and high bitrate lossy compression, and this is the reason that some prefer FLACs etc
Depends of how you define "hear the difference". If you gave me two files encoded at 320 with the Fraunhofer codec, and one lossless file, and unlimited time, I could tell the lossless one from the others. Granted Frau isn't quite modern, but its certainly possible
Depends on how good the sound engineers are. A lot can be gained by higher resolution and sample rate in the mastering stage, but by using a good low pass filter and dithering (and dithering is not really necessary, http://developers.slashdot.org/story/13/02/27/1547244/xiph-episode-2-digital-show-tell ) basically all audible information is captured in 44.1kHz / 16. Your speakers probably don't go much above 20 kHz anyway, so anything beyond 44.1kHz will only cause distortion (aliasing), see post by MetalliQaZ "Debunked" below.
Of course we needed Schmidt to tell us that. It's easy to say we "knew" these things in hindsight. It would be like saying we didn't need the LHC because we knew the Higgs was there.
Speed does mean something. My 2 year old Nexus S has frequent delays of 5-10 seconds. It seems slower than when I bought it, but I can't see any reason why Google would make their OS slower, so I have no idea why it's like that. If there is a delay of order 100 ms, it may be caused by frequency scaling and other power saving features, but multiple second delays would definitely be fixed by a faster CPU. There is also RAM, though. Hard to tell if it's swapping or running at full speed when it's all solid state.
IMHO this is more like a work-around that a real solution. You have to allocate RAM for each application VM, which is really unpractical and will impact performance. A better approach seems to be to build security from the top down by extending chroot into things like Linux Containers and FreeBSD jails. You get similar isolation for applications, but full system performance. You don't have the extra security of isolation for the drivers, that wouldn't be possible in Linux without virtualisation. However, in Qubes you still need some protocol for the other parts of the system to talk to the network VM, so if someone owns the network VM, they can quite possible also get at the other parts of the system (admittedly much more difficult than the instant root (or better) that can be obtained by owning a driver in standard Linux).
Qubes OS has created a nice innovation on the main X11 desktop. Applications from different security domains are clearly marked, so that it is difficult to fool the user into entering information in the wrong place. X11 is usually a bottleneck when it comes to isolation, and Qubes fixes this.
Since when is it against the law to modify an Xbox?
Haven't you seen the stories about phone unlocking recently? This is exactly the same thing. Once some hardware product involves DRM and software, all bets are off.
The Xbox DRM is designed to control exactly what software runs on the Xbox, so you are by necessity breaking it if you modify the Xbox. The anti-circumvention rule in the DMCA makes it illegal to break DRM, regardless of whether what you accomplish by breaking it would otherwise be legal under rules such as fair use.
It would be hard to even argue that the software on the phone is copied in any way when unlocking a phone, yet people don't seem to question if the anti-circumvention rule applies.
Will gcc use AVX or FMA3 if I write normal code in C++? How about Java and Python / numpy, could it be that python actually gets faster than C++ if gcc doesn't take advantage of these technologies?
Actual error rates in good memory are very low. I didn't see a single error for a year. The main benefit of ECC on workstations is to detect memory that is slightly bad, which passes hours of memtest86, but still gives you errors maybe every month on your workload. This requires you to monitor ECC errors, and get alerts, so you can replace the DIMMs that give errors repeatedly. The problem is that ECC monitoring for the new Intel chips is not available in Linux (as far as I can tell, cheeky plug for my stackexchange question http://unix.stackexchange.com/questions/67999/how-to-monitor-ram-ecc-errors-on-intel-processor-in-linux )
This is a purely defensive installation, and the cost isn't huge by military proportions. No need to even name an enemy. In fact, they could probably install three layers of this system for $3 billion and have 87 % success rate (if the probability of success does not depend on the trajectory). The chance for a nuclear strike is small, but nukes are a "holy grail" for all small totalitarian regimes, of which there are a couple in east Asia, and it's better to not be taken by surprise.
It's a brilliant, almost tit for tat response to actual trolls who do it just to spread evil and waste other people's time. It is pretty devastating if it's used for other commenters, who maybe just have a provocative writing style or unpopular views. I hope it's being used with great caution
He fails to correct for the fact that he may have changed in 8 years. If he was 8 years older in 2005 he may have wanted fewer gadgets then. He admits it himself with the consoles. I don't usually like to argue based on age, but it seems like a glaring omission in the analysis. To his credit, he hasn't changed that much physically.
If these lights or their controllers are publicly adressable on the Internet, they will be hacked. Fortunately the tech is still in its infancy and the people who install these things probably know how to maintain and update them. The damage that a hacked light or a central heater can do mostly amounts to an annoyance and increased power use, assuming that it has proper hardware protections and manual overrides.
Another safety issue is if burglars are armed with RF jammers; they could prevent the house owner from turning on the lights and even calling the emergency number. This is not as bad as it seems though, because burglars prefer empty houses and many are junkies and may not even be armed with proper clothes, much less a jammer
the outdated Unix security model, which assumes that the program is the user and would do roughly what the user wanted (maybe true in the 1970s on shared university systems, but obvious nonsense now).
It's a good thing that it evolved this way, because insecurity also makes it easier for the programmer. If malware and cyber* was as rampant in the late 1990s as it is now, we would have some horrible locked down computers which only did ~6 things that were blessed by the manufacturer. Today is a good time to start making systems more secure, but there also needs to be an open-ended environment where small programs can share data without any restrictions.
Warning: progressively more off topic rant. Which ROM was it that worked for you? I'm about to give up on Android because of privacy issues, but there aren't many credible alternatives. I think iPhone actually has much better privacy, at least if iCloud sync can be disabled, but the open/closed difference is more important for me than privacy for now (even after this story, which only affects the Google store). On the new version of Android I can't even add local contacts. I'd have to set up an Exchange server if I didn't want to share my contacts with a US company (I have nothing against US companies vs. other companies, just pointing out the limited choice). [I just thought of an app idea, to make a "Local PseudoAccount" app that acts like an online account, but stores everything in an encrypted SQLite file on the SD card. It would only solve one minor privatcy problem though]
Sometimes I feel that I am now officially "old" (at the age of 26) for caring about this stuff. The next wave of technology may be to use accounts in huge centralised databases for all computing needs. If we're going to have a sci-fi future, that may indeed be the way to go, and in sci-fi they don't usually address who runs the infrastructure and has access to the data. For this reason I have considered to just give up and spew my data over all kinds of services and use Facebook connect, etc,etc.
The best way to use Linux is to stay five years behind the development. RHEL (and derivatives, like Scientific Linux and CentOS) still use Gnome 2, it uses SysV init scripts, and everything is stable and lovely. This is what would happen for Windows and Apple too, but the development happens internally, whereas for Linux is it visible for all.
That's not to say this should be ignored, I find it very interesting and I'll consider watching the hour and a half long video in the comment above me. Cutting-edge distros are however only useful for enthusiasts, and to get real work done you need something more stable (this is not strictly true, there are many kinds of work you can do with just a web browser, but I hope you get what I'm saying). I'll be very interested to see what things get included in RHEL as time goes on. Will they include only the good stuff, or will they add everything... So in 2019 I may indeed switch to Wayland, if its' any good.
The anit-circumvention rule has gone off the rails. One thing is breaking e-book DRM for the blind to read them, or ripping DVDs for backup. Then you are actually copying something.
Now phone unlocking is different: you are not actually copying *anything*. This should not be a copyright issue at all. It's one thing if they want to make a rule to let a manufacturer limit what you can do with a product you bought, for their business interests. Try to get that law to pass, etc. I'll hate you, but less. Sneaking it in through copyright, however, is just dishonest and disgusting. It also demonstrates how the world of copyright is much more restrictive than the laws that govern the physical world.
I also wonder if I took the ROM and flash chips out of an iPhone and overwrote the content with my own OS, if that would still be legal. Is the tenuous copyright argument tied to the OS & software, or do they not even bother trying to point to a copyrighted work?
I now realise that my post was a bit off topic for this thread. I sometimes turn off Flash in the Firefox add-on manager, and turn it back on when I need it for a video, but the options posted in this thread seem much better.
I figure there are lots of alternatives to sites with obnoxious ads. I stopped going to dictionary.com when they played an ad with sound, and I'm using merriam-webster.com instead (they have a rate limit on the searches, but I usually just need one). When wikipedia showed the huge donation banner that popped up and down while I was reading, I significantly reduced my use of Wikipedia. I have to admit I still go there quite a lot, but when there's another plausible link in the google results, e.g. Wolfram MathWorld, I usually click that instead. The only ads I mind are the ones that force me to interact with them (or if the frantically blink, etc.). I find about 0.001 % of the ads interesting enough to look closer at, and about half of them worthy of a click, so when an ad forces itself on me, there's a probability of ~1 of wasting my time. [When there's a full page ad with an option to "Skip this ad" I usually just look at something else instead of skipping it, I just don't want to interact with them]
It's not unthinkable that real people would mod this down. It's fine to point out that it's an advertisement, but when the first thread just keeps going on and on about it, I could see how some people would consider that to be off topic.
Slashdot Mirror
Very cool resource. I didn't know that lossy codecs relied on an "uncertainty principle" like relation, but it seems likely that hearing is not linear. I don't think that anyone seriously claimed that it was *impossible* to tell the difference between lossless and high bitrate lossy compression, and this is the reason that some prefer FLACs etc
Depends of how you define "hear the difference". If you gave me two files encoded at 320 with the Fraunhofer codec, and one lossless file, and unlimited time, I could tell the lossless one from the others. Granted Frau isn't quite modern, but its certainly possible
Depends on how good the sound engineers are. A lot can be gained by higher resolution and sample rate in the mastering stage, but by using a good low pass filter and dithering (and dithering is not really necessary, http://developers.slashdot.org/story/13/02/27/1547244/xiph-episode-2-digital-show-tell ) basically all audible information is captured in 44.1kHz / 16. Your speakers probably don't go much above 20 kHz anyway, so anything beyond 44.1kHz will only cause distortion (aliasing), see post by MetalliQaZ "Debunked" below.
I really wanted to mod this up because it's a great resource.
But: it's actually off topic. The question here is about using psychoacoustic compression like MP3 and AAC, not about sample rate or bit depth.
Auto-play audio ad => instant close
Of course we needed Schmidt to tell us that. It's easy to say we "knew" these things in hindsight. It would be like saying we didn't need the LHC because we knew the Higgs was there.
Speed does mean something. My 2 year old Nexus S has frequent delays of 5-10 seconds. It seems slower than when I bought it, but I can't see any reason why Google would make their OS slower, so I have no idea why it's like that. If there is a delay of order 100 ms, it may be caused by frequency scaling and other power saving features, but multiple second delays would definitely be fixed by a faster CPU. There is also RAM, though. Hard to tell if it's swapping or running at full speed when it's all solid state.
IMHO this is more like a work-around that a real solution. You have to allocate RAM for each application VM, which is really unpractical and will impact performance. A better approach seems to be to build security from the top down by extending chroot into things like Linux Containers and FreeBSD jails. You get similar isolation for applications, but full system performance. You don't have the extra security of isolation for the drivers, that wouldn't be possible in Linux without virtualisation. However, in Qubes you still need some protocol for the other parts of the system to talk to the network VM, so if someone owns the network VM, they can quite possible also get at the other parts of the system (admittedly much more difficult than the instant root (or better) that can be obtained by owning a driver in standard Linux).
Qubes OS has created a nice innovation on the main X11 desktop. Applications from different security domains are clearly marked, so that it is difficult to fool the user into entering information in the wrong place. X11 is usually a bottleneck when it comes to isolation, and Qubes fixes this.
Since when is it against the law to modify an Xbox?
Haven't you seen the stories about phone unlocking recently? This is exactly the same thing. Once some hardware product involves DRM and software, all bets are off.
The Xbox DRM is designed to control exactly what software runs on the Xbox, so you are by necessity breaking it if you modify the Xbox. The anti-circumvention rule in the DMCA makes it illegal to break DRM, regardless of whether what you accomplish by breaking it would otherwise be legal under rules such as fair use.
It would be hard to even argue that the software on the phone is copied in any way when unlocking a phone, yet people don't seem to question if the anti-circumvention rule applies.
Will gcc use AVX or FMA3 if I write normal code in C++? How about Java and Python / numpy, could it be that python actually gets faster than C++ if gcc doesn't take advantage of these technologies?
Actual error rates in good memory are very low. I didn't see a single error for a year. The main benefit of ECC on workstations is to detect memory that is slightly bad, which passes hours of memtest86, but still gives you errors maybe every month on your workload. This requires you to monitor ECC errors, and get alerts, so you can replace the DIMMs that give errors repeatedly. The problem is that ECC monitoring for the new Intel chips is not available in Linux (as far as I can tell, cheeky plug for my stackexchange question http://unix.stackexchange.com/questions/67999/how-to-monitor-ram-ecc-errors-on-intel-processor-in-linux )
This is a purely defensive installation, and the cost isn't huge by military proportions. No need to even name an enemy. In fact, they could probably install three layers of this system for $3 billion and have 87 % success rate (if the probability of success does not depend on the trajectory). The chance for a nuclear strike is small, but nukes are a "holy grail" for all small totalitarian regimes, of which there are a couple in east Asia, and it's better to not be taken by surprise.
It's going to be like hacker/cracker, isn't it :(
It's a brilliant, almost tit for tat response to actual trolls who do it just to spread evil and waste other people's time. It is pretty devastating if it's used for other commenters, who maybe just have a provocative writing style or unpopular views. I hope it's being used with great caution
He fails to correct for the fact that he may have changed in 8 years. If he was 8 years older in 2005 he may have wanted fewer gadgets then. He admits it himself with the consoles. I don't usually like to argue based on age, but it seems like a glaring omission in the analysis. To his credit, he hasn't changed that much physically.
If these lights or their controllers are publicly adressable on the Internet, they will be hacked. Fortunately the tech is still in its infancy and the people who install these things probably know how to maintain and update them. The damage that a hacked light or a central heater can do mostly amounts to an annoyance and increased power use, assuming that it has proper hardware protections and manual overrides.
Another safety issue is if burglars are armed with RF jammers; they could prevent the house owner from turning on the lights and even calling the emergency number. This is not as bad as it seems though, because burglars prefer empty houses and many are junkies and may not even be armed with proper clothes, much less a jammer
the outdated Unix security model, which assumes that the program is the user and would do roughly what the user wanted (maybe true in the 1970s on shared university systems, but obvious nonsense now).
It's a good thing that it evolved this way, because insecurity also makes it easier for the programmer. If malware and cyber* was as rampant in the late 1990s as it is now, we would have some horrible locked down computers which only did ~6 things that were blessed by the manufacturer. Today is a good time to start making systems more secure, but there also needs to be an open-ended environment where small programs can share data without any restrictions.
Sometimes I feel that I am now officially "old"
Ten minutes later I got an ad for "Avoid running out of money during retirement" .. I wonder if Google is getting that good ;)
Warning: progressively more off topic rant.
Which ROM was it that worked for you? I'm about to give up on Android because of privacy issues, but there aren't many credible alternatives. I think iPhone actually has much better privacy, at least if iCloud sync can be disabled, but the open/closed difference is more important for me than privacy for now (even after this story, which only affects the Google store). On the new version of Android I can't even add local contacts. I'd have to set up an Exchange server if I didn't want to share my contacts with a US company (I have nothing against US companies vs. other companies, just pointing out the limited choice). [I just thought of an app idea, to make a "Local PseudoAccount" app that acts like an online account, but stores everything in an encrypted SQLite file on the SD card. It would only solve one minor privatcy problem though]
Sometimes I feel that I am now officially "old" (at the age of 26) for caring about this stuff. The next wave of technology may be to use accounts in huge centralised databases for all computing needs. If we're going to have a sci-fi future, that may indeed be the way to go, and in sci-fi they don't usually address who runs the infrastructure and has access to the data. For this reason I have considered to just give up and spew my data over all kinds of services and use Facebook connect, etc,etc.
Depends on how you define "success". If you define it as being popular (i.e. has many users), the thesis in the article is basically tautological.
The best way to use Linux is to stay five years behind the development. RHEL (and derivatives, like Scientific Linux and CentOS) still use Gnome 2, it uses SysV init scripts, and everything is stable and lovely. This is what would happen for Windows and Apple too, but the development happens internally, whereas for Linux is it visible for all.
That's not to say this should be ignored, I find it very interesting and I'll consider watching the hour and a half long video in the comment above me. Cutting-edge distros are however only useful for enthusiasts, and to get real work done you need something more stable (this is not strictly true, there are many kinds of work you can do with just a web browser, but I hope you get what I'm saying). I'll be very interested to see what things get included in RHEL as time goes on. Will they include only the good stuff, or will they add everything... So in 2019 I may indeed switch to Wayland, if its' any good.
The anit-circumvention rule has gone off the rails. One thing is breaking e-book DRM for the blind to read them, or ripping DVDs for backup. Then you are actually copying something.
Now phone unlocking is different: you are not actually copying *anything*. This should not be a copyright issue at all. It's one thing if they want to make a rule to let a manufacturer limit what you can do with a product you bought, for their business interests. Try to get that law to pass, etc. I'll hate you, but less. Sneaking it in through copyright, however, is just dishonest and disgusting. It also demonstrates how the world of copyright is much more restrictive than the laws that govern the physical world.
I also wonder if I took the ROM and flash chips out of an iPhone and overwrote the content with my own OS, if that would still be legal. Is the tenuous copyright argument tied to the OS & software, or do they not even bother trying to point to a copyrighted work?
I now realise that my post was a bit off topic for this thread. I sometimes turn off Flash in the Firefox add-on manager, and turn it back on when I need it for a video, but the options posted in this thread seem much better.
I figure there are lots of alternatives to sites with obnoxious ads. I stopped going to dictionary.com when they played an ad with sound, and I'm using merriam-webster.com instead (they have a rate limit on the searches, but I usually just need one). When wikipedia showed the huge donation banner that popped up and down while I was reading, I significantly reduced my use of Wikipedia. I have to admit I still go there quite a lot, but when there's another plausible link in the google results, e.g. Wolfram MathWorld, I usually click that instead. The only ads I mind are the ones that force me to interact with them (or if the frantically blink, etc.). I find about 0.001 % of the ads interesting enough to look closer at, and about half of them worthy of a click, so when an ad forces itself on me, there's a probability of ~1 of wasting my time. [When there's a full page ad with an option to "Skip this ad" I usually just look at something else instead of skipping it, I just don't want to interact with them]
It's not unthinkable that real people would mod this down. It's fine to point out that it's an advertisement, but when the first thread just keeps going on and on about it, I could see how some people would consider that to be off topic.