Weird. Digital recording and mastering was definitely a thing at that time. You’d think they would have been crashing the HDDs of PCs in the recording studios.
Not weird at all. This problem manifested only with some model of 5400RPM laptop hard drive (2.5"), but a recording studio would likely have been using 7200RPM 3.5" desktop drives. Different resonant frequencies, more sturdy mounting, more distance between the speakers and the hard drives.
/* Emits a 7-Hz tone for 10 seconds.
True story: 7 Hz is the resonant
frequency of a chicken's skull cavity.
This was determined empirically in
Australia, where a new factory
generating 7-Hz tones was located too
close to a chicken ranch: When the
factory started up, all the chickens
died.
Your PC may not be able to emit a 7-Hz tone. */
#include
int main(void)
{
sound(7);
delay(10000);
nosound();
return 0;
}
This is someone retelling a story they were told by a co-worker of an event over 20 years prior. It’s not surprising that he doesn’t go into the details of exactly what was tried, beyond the key parts of the story.
Not an expert here, so I’m genuinely curious how could a video stream (edit: with muted audio stream) possibly cause another laptop in close proximity to crash?
What is claimed in TFA is that the hard drive resonate frequency reacts to the Janet Jackson video in bad ways because that music video puts out music that interferes with what the hard drive expects.
TFA was lacking details so this is merely a retelling.
Obviously not the video but the accompanying audio track. Could also just be a made up apocryphal engineering story that never actually happened exactly as described. Engineering as a profession is chock full of them but they do tend to be memorable parables of things to keep in mind when working on a relevant piece of tech.
What is definitely well documented is Brendan Gregg’s related discovery of performance degradation in servers from vibration of sibling servers / clapping nearby that caused spinning disks to pause their heads.
I doubt it could, but when you run into a problem that defies your understanding of reality, you might try out responses that also defy your understanding of reality, in the hopes you might gain the missing insight somewhere along the way, yeah?
If this is just a fiction novel world‑building question: The video pixels create a bitstream to bitbang the gpu bus into emitting a 2.4‑gigahertz EMF signal to exploit a flaw in the Wi‑Fi driver.
Also not an expert, it would have to be EMI or maybe the bright light was causing LEDs on the nearby laptop to generate voltage. LEDs can poorly work in reverse.
I'd love to know whether that story is actually true.
Some dude hears somebody tell a story about sth 20 years ago, puts it in a blog, and here we are on HN, nobody questioning whether it's actually accurate. Of course Raymond Chen isn't just any random person, but the more important it would be to actually check? I mean, who hasn't heard people tell stories from decades ago, including colleagues reminiscing about the good old times "before y'all were born" only to realize later that it was vastly exaggerated or even outright made up.
Anybody around here with some actual first-hand info or at least another source besides this blog entry? I'd love to hear!
It's like Mark Twain and the rules for reselling a slave in Missouri https://medium.com/p/fe48ea07ad20
"the free black man in Missouri could only remain in the state for 6 months before being taken and put on auction as a slave." only it turned out to be false, and evidently made up by Twain for reasons of fiction.
Never let the truth get in the way of a good story. That's my motto. Now let me tell you about the time that we dug up this dinosaur egg and hatched it.
I believe it because it's a plausible variant of what I call the "Fus Ro Data Loss" vulnerability: shouting at hard drives causes them to resonate in a way that affects their ability to access data.
Technically, that magnetic spinning HDD can work even after decades if maintained safely (no dust, no extreme heat) and without stress, even if it is not switched on for years.
In fact, if a magnetic HDD crashes, you may still recover some or all of the data by doing something hardcore, such as letting it sit for some hours in the freezer of your refrigerator, or immersing it in a bowl of rice overnight.
However, SSDs (and other flash storage devices) need to be switched on once in few months, otherwise there's a chance that some data stored in them may be permanently lost, as some cells may loose their power.
I feel like maybe you didn't understand the meaning of that last bit you quoted from Tom's Hardware. To be clear: the standard for consumer SSDs is 1 year of unpowered data retention after the drive's full write endurance rating has been exhausted.
The experiment Tom's is reporting on found twelve instances of data corruption on a low-end drive that had been subjected to over two thousand full drive writes, four times its rated write endurance, then left on a shelf for two years. This is a demonstration of a bottom of the barrel SSD wildly exceeding expectations.
It's really important in conversations like this to accurately convey not just the existence of the failure mode, but also the realistic chances of running into this problem, and the extent of the problem when it does manifest. If a deliberate torture test can only produce a few kilobytes of data corruption after twice the duration and four times the abuse the drive is supposed to be able to handle, this problem should be described as extremely minor.
But the materials on the CD eventually break down, sometimes as soon as within 5 years. So you can look into MDisc, which purports 100 years…but only in theory since the tests are just approximations of what would actually happen.
The claim you're responding to is that hard drives lose "magnetic charge" at a rate of 1% per year, not that bits get corrupted at a rate of 1% per year. The error correction in hard drives is far simpler and weaker than what's used in SSDs, but it does exist. So we should expect that there's a significant margin for data degradation before any observable data corruption begins. (This is true for SSDs, too; the first symptom of data degradation is reduced read performance as slower, more complex error correction methods kick in, then much later the host starts to actually get read errors or bad data.)
The magnetic strength of particles on the disk can decay at 1% per year, but the drive won't have issues reading it until it falls below a threshold where it can no longer be read. It could take decades.
We still talk about "bugs" (99+% of computer defects in the past 70+ years have not been caused by insects) and "AJAX" (long after most of these requests use JSON instead of XML).
> Yes, I know which “major computer manufacturer” it is, and no, I’m not telling. This is consistent with longstanding blog policy that companies are not identified in stories, because the point of the story to teach something, not to call out companies for derision.
He might not since it comes via a friend. Or he's forgotten since.
Also seems not unreasonable for an employee like him not to specifically name and shame hardware partners. Maybe it'd all be fine, but I wouldn't blame him at all for not wanting to risk it.
He genuinely might not know. I worked on a similar incident when our video encoder caused about 30% of a pretty mainstream mobile handset to hard lock when recieving a stream, requiring the battery to be removed to reboot the device.
Neither us nor the OEM ever figured out why. They suspected that it was a weird combination of different bin combinations from different parts, but ultimately we had to change the method of delivering video to stop it happening.
The Dutch broadcasting service hired me to figure out why their homepage was crashing browsers. I turned out to be an animated GIF of two speakers that had an extra 0 interval frame in it which caused IE to crash... it doesn't take much.
Very funny, reminds me of how Jennifer Lopez created Google Image Search when she wore a very deep cut green dress in 2000. So many people searched for "Jennifer Lopez Green Dress" that the search team realized they needed to include images in the search results. https://www.project-syndicate.org/magazine/google-european-c... https://www.thecut.com/2019/09/jennifer-lopez-walks-in-versa...
Weird. Digital recording and mastering was definitely a thing at that time. You’d think they would have been crashing the HDDs of PCs in the recording studios.
Not weird at all. This problem manifested only with some model of 5400RPM laptop hard drive (2.5"), but a recording studio would likely have been using 7200RPM 3.5" desktop drives. Different resonant frequencies, more sturdy mounting, more distance between the speakers and the hard drives.
https://everything2.com/title/7+hertz+-+the+resonant+frequen...
Example (for both functions):
from the comments over there (2002)Not very likely
https://skeptics.stackexchange.com/a/54400
Extra points for people who avoid gratuitous clickbait
Just a small nitpick: the Tacoma Narrows bridge didn’t collapse because of resonance but because of flutter. It’s a common misconception.
For resonance the external driving force must match the resonance frequency of the system, but wind is rarely/never purely sinusoidal.
Is flutter a derivative like jerk?
https://en.wikipedia.org/wiki/Aeroelasticity#Flutter
Related. Others?
Janet Jackson had the power to crash laptop computers (2022) - https://news.ycombinator.com/item?id=41534483 - Sept 2024 (79 comments)
Janet Jackson had the power to crash laptop computers - https://news.ycombinator.com/item?id=32483211 - Aug 2022 (12 comments)
https://youtu.be/tDacjrSCeq4 reminds me of this gem.
I was reminded of this MythBusters episode: Tesla's Earthquake Machine
https://www.youtube.com/watch?v=LHsHiKtjoag
Similary story from Apple: https://youtu.be/C5d151lqJsA?t=108
That’s a heck of debugging.
Why didn't they mute the volume to see if it was the video or audio stream causing problems?
This is someone retelling a story they were told by a co-worker of an event over 20 years prior. It’s not surprising that he doesn’t go into the details of exactly what was tried, beyond the key parts of the story.
Not an expert here, so I’m genuinely curious how could a video stream (edit: with muted audio stream) possibly cause another laptop in close proximity to crash?
What is claimed in TFA is that the hard drive resonate frequency reacts to the Janet Jackson video in bad ways because that music video puts out music that interferes with what the hard drive expects.
TFA was lacking details so this is merely a retelling.
Obviously not the video but the accompanying audio track. Could also just be a made up apocryphal engineering story that never actually happened exactly as described. Engineering as a profession is chock full of them but they do tend to be memorable parables of things to keep in mind when working on a relevant piece of tech.
What is definitely well documented is Brendan Gregg’s related discovery of performance degradation in servers from vibration of sibling servers / clapping nearby that caused spinning disks to pause their heads.
I doubt it could, but when you run into a problem that defies your understanding of reality, you might try out responses that also defy your understanding of reality, in the hopes you might gain the missing insight somewhere along the way, yeah?
If this is just a fiction novel world‑building question: The video pixels create a bitstream to bitbang the gpu bus into emitting a 2.4‑gigahertz EMF signal to exploit a flaw in the Wi‑Fi driver.
Also not an expert, it would have to be EMI or maybe the bright light was causing LEDs on the nearby laptop to generate voltage. LEDs can poorly work in reverse.
That's Miss Jackson if you're Windows.
I'd love to know whether that story is actually true.
Some dude hears somebody tell a story about sth 20 years ago, puts it in a blog, and here we are on HN, nobody questioning whether it's actually accurate. Of course Raymond Chen isn't just any random person, but the more important it would be to actually check? I mean, who hasn't heard people tell stories from decades ago, including colleagues reminiscing about the good old times "before y'all were born" only to realize later that it was vastly exaggerated or even outright made up.
Anybody around here with some actual first-hand info or at least another source besides this blog entry? I'd love to hear!
It's like Mark Twain and the rules for reselling a slave in Missouri https://medium.com/p/fe48ea07ad20 "the free black man in Missouri could only remain in the state for 6 months before being taken and put on auction as a slave." only it turned out to be false, and evidently made up by Twain for reasons of fiction.
Never let the truth get in the way of a good story. That's my motto. Now let me tell you about the time that we dug up this dinosaur egg and hatched it.
I believe it because it's a plausible variant of what I call the "Fus Ro Data Loss" vulnerability: shouting at hard drives causes them to resonate in a way that affects their ability to access data.
https://m.youtube.com/watch?v=tDacjrSCeq4
Computers are so weird.
Analog devices pretending to be digital
Another reason to step away from spinning rust.
Thank dog for SSDs
Technically, that magnetic spinning HDD can work even after decades if maintained safely (no dust, no extreme heat) and without stress, even if it is not switched on for years.
In fact, if a magnetic HDD crashes, you may still recover some or all of the data by doing something hardcore, such as letting it sit for some hours in the freezer of your refrigerator, or immersing it in a bowl of rice overnight.
However, SSDs (and other flash storage devices) need to be switched on once in few months, otherwise there's a chance that some data stored in them may be permanently lost, as some cells may loose their power.
"As a reminder, an SSD's endurance rating is calculated based on how long it can store data if left unplugged after a certain amount of data has been written": https://www.tomshardware.com/pc-components/storage/unpowered...
I feel like maybe you didn't understand the meaning of that last bit you quoted from Tom's Hardware. To be clear: the standard for consumer SSDs is 1 year of unpowered data retention after the drive's full write endurance rating has been exhausted.
The experiment Tom's is reporting on found twelve instances of data corruption on a low-end drive that had been subjected to over two thousand full drive writes, four times its rated write endurance, then left on a shelf for two years. This is a demonstration of a bottom of the barrel SSD wildly exceeding expectations.
It's really important in conversations like this to accurately convey not just the existence of the failure mode, but also the realistic chances of running into this problem, and the extent of the problem when it does manifest. If a deliberate torture test can only produce a few kilobytes of data corruption after twice the duration and four times the abuse the drive is supposed to be able to handle, this problem should be described as extremely minor.
HDDs also lose magnetic charge over time, about 1% per year. So you need to periodically spin up and rewrite the data every few years.
CD drives however, can store data indefinitely without needing refreshing.
But the materials on the CD eventually break down, sometimes as soon as within 5 years. So you can look into MDisc, which purports 100 years…but only in theory since the tests are just approximations of what would actually happen.
Totally could be a defect in the lamination we don’t find out for years yet
CD-R media is of limited shelf life as well though
Having had drives which sat for many years and spun right back up without corruption makes me think 1% is too generous maybe 0.05% per year at most
The claim you're responding to is that hard drives lose "magnetic charge" at a rate of 1% per year, not that bits get corrupted at a rate of 1% per year. The error correction in hard drives is far simpler and weaker than what's used in SSDs, but it does exist. So we should expect that there's a significant margin for data degradation before any observable data corruption begins. (This is true for SSDs, too; the first symptom of data degradation is reduced read performance as slower, more complex error correction methods kick in, then much later the host starts to actually get read errors or bad data.)
The magnetic strength of particles on the disk can decay at 1% per year, but the drive won't have issues reading it until it falls below a threshold where it can no longer be read. It could take decades.
Spinning rust may be defeated by Janet Jackson, but your chip storage is defeated by just sitting undisturbed in a drawer or a closet for too long...
Platters are not made of iron (or even steel) and neither is the surface, so I’m not sure why rust comes into the picture.
The platters used to be coated with iron oxide as the magnetic coating.
Modern ones use more exotic materials.
Yes but that was like 40 years ago. I think by now the phraseology should reflect reality after about forty years.
Ironically, I only started hearing the term being used in the past decade, as a colloquial alternative to solid-state storage.
We still talk about "bugs" (99+% of computer defects in the past 70+ years have not been caused by insects) and "AJAX" (long after most of these requests use JSON instead of XML).
“Phraseology” is subtly different.
You mean “vocabulary”, “terminology”, possibly “nomenclature”.
How many horsepower does your car have?
I love this story! Ha
> certain models
Why the weasel words? Does Raymond Chen not know which models? Or is it actually apocryphal.
> Yes, I know which “major computer manufacturer” it is, and no, I’m not telling. This is consistent with longstanding blog policy that companies are not identified in stories, because the point of the story to teach something, not to call out companies for derision.
From the follow-up post: https://devblogs.microsoft.com/oldnewthing/20220920-00/?p=10...
He might not since it comes via a friend. Or he's forgotten since.
Also seems not unreasonable for an employee like him not to specifically name and shame hardware partners. Maybe it'd all be fine, but I wouldn't blame him at all for not wanting to risk it.
In general, he never names companies in his posts if there is anything potentially negative in there.
He genuinely might not know. I worked on a similar incident when our video encoder caused about 30% of a pretty mainstream mobile handset to hard lock when recieving a stream, requiring the battery to be removed to reboot the device.
Neither us nor the OEM ever figured out why. They suspected that it was a weird combination of different bin combinations from different parts, but ultimately we had to change the method of delivering video to stop it happening.
The Dutch broadcasting service hired me to figure out why their homepage was crashing browsers. I turned out to be an animated GIF of two speakers that had an extra 0 interval frame in it which caused IE to crash... it doesn't take much.