Re:Debugging Behind the Iron Curtain (2010)
by mav3r1ck on Tuesday March 20, 2018 @12:03PM
(#16629757)
Wow. I’ve discovered many bugs in my life. But none caused me to fear for my life.
>There were often significant food shortages in the Soviet Union, and the government plan was to mix the meat from Chernobyl-area cattle with the uncontaminated meat from the rest of the country. This would lower the average radiation levels of the meat without wasting valuable resources. Upon discovering this, Sergei immediately filed immigration papers with any country that would listen. The computer crashes resolved themselves as radiation levels dropped over time.
> It was involved in at least six accidents between 1985 and 1987, in which patients were given massive overdoses of radiation. Because of concurrent programming errors, it sometimes gave its patients radiation doses that were hundreds of times greater than normal, resulting in death or serious injury.
Re:Debugging Behind the Iron Curtain (2010)
by jzymbaluk on Tuesday March 20, 2018 @11:20AM
(#16629303)
I love reading these bug-hunting war stories, there's no breakpoint for radioactive cows in GDB! Another favorite of mine is the 500-mile email bug[1], if there's a bigger collection of these stories anywhere on the internet, I'd love to read them
Re:Debugging Behind the Iron Curtain (2010)
by planteen on Tuesday March 20, 2018 @04:16PM
(#16632140)
I kind of doubt this story. What kind of radiation would you have from Chernobyl?
My understanding is that bit flips (SEEs) are typically caused by energetic particles like cosmic rays. To simulate upset damage on Earth, you typically need to decap the IC and go to a testing facility that directly hits the wafer with particles. It seems unlikely that somewhere near Chernobyl that there would be particles energetic enough that didn't get stopped by the IC packaging.
Gamma radiation gives total ionizing dose effects that degrade chips until the point they no longer work. Those shoot right through the chip packaging. But I don't think it typically causes upsets.
Re:Debugging Behind the Iron Curtain (2010)
by virtuexru on Tuesday March 20, 2018 @11:05AM
(#16629162)
Having issues loading the text on this page. Seems to keep "blinking" out for me. Is there anyway we can get a printer friendly or text only version? Super interested in this article.
Re:Debugging Behind the Iron Curtain (2010)
by mtve on Wednesday March 21, 2018 @10:07AM
(#16638073)
1. anonymous Sergey as a story source, without any verificable identity.
2. "he went drinking with a few military personnel", so it's retelling of rehearsal, even more trust.
3. "government plan was to mix the meat from Chernobyl-area cattle with the uncontaminated meat", that's seems plausible. contamination of meat was "up to 1,0*10-6 Ci/kg" (according to widely cited secret "annex 10 of protocol #32"), it's 300 times of banana equivalent.
from the same scary stories fansites, there were 34 Кtons of bad meat (it's one thousand of railway carriages btw). so not so dramatic, definitely not a near nuclear blast.
4. [computer was] "located in a building close to the railroad tracks". inverse-square law, walls, exposure time from a moving train etc. i would more believe in vibration caused by trains as a cause of problems.
5. the SM-1800 computer:
KR580VM80A cpu (8080 clone) is 6 mkm and 2 MHz (compare to modern 10 nm and 2 GHz),
K565RU3A ram (4116 analog, like in ZX Spectrum), and other components. even the computer case was solid metal.
i can not tell you precisely, but there should be a powerful source of radiation to crash such computer reproducibly.
so no, i don't buy it. sorry for hurting your radio- and russo-fobias.
Re:Debugging Behind the Iron Curtain (2010)
by abritinthebay on Wednesday March 21, 2018 @03:36PM
(#16641539)
> sorry for hurting your radio- and russo-fobias.
That's a mighty big chip you have on your shoulder there. Sorry to disappoint your ego but I was curious as to your objections. They're... not very good.
The inverse square law one is the worst - you make tons of assumptions about the building and then screw up the basic science. A few inches of lead is enough to block gamma rays -- which is why you wear a lead apron when you get x-rays -- but at least 3 meters of concrete are needed to stop them.
Assuming a standard brick or breeze block building and normal computer metal case plus a meter or two of separation (common at railways in Europe at least, when stuff is going by slowly) that wouldn't stop much gamma radiation at all.
Other than "it's a semi-anonymous story" you really don't have much in the way of objection here.
Re:Debugging Behind the Iron Curtain (2010)
by planteen on Tuesday March 20, 2018 @04:20PM
(#16632182)
Odds are the camera flash discharge caused electromagnetic interference (EMI) that disturbed your electronics rather than the visible light and UV radiation from the flash itself. Wafers are photo sensitive, but light is blocked by the packaging. Did you have decapped parts?
>There were often significant food shortages in the Soviet Union, and the government plan was to mix the meat from Chernobyl-area cattle with the uncontaminated meat from the rest of the country. This would lower the average radiation levels of the meat without wasting valuable resources. Upon discovering this, Sergei immediately filed immigration papers with any country that would listen. The computer crashes resolved themselves as radiation levels dropped over time.
A race condition which killed people: https://en.wikipedia.org/wiki/Therac-25
> It was involved in at least six accidents between 1985 and 1987, in which patients were given massive overdoses of radiation. Because of concurrent programming errors, it sometimes gave its patients radiation doses that were hundreds of times greater than normal, resulting in death or serious injury.
[1] http://web.mit.edu/jemorris/humor/500-miles
My understanding is that bit flips (SEEs) are typically caused by energetic particles like cosmic rays. To simulate upset damage on Earth, you typically need to decap the IC and go to a testing facility that directly hits the wafer with particles. It seems unlikely that somewhere near Chernobyl that there would be particles energetic enough that didn't get stopped by the IC packaging.
Gamma radiation gives total ionizing dose effects that degrade chips until the point they no longer work. Those shoot right through the chip packaging. But I don't think it typically causes upsets.
https://www.isis.stfc.ac.uk/Pages/ChipIR.aspx
2. "he went drinking with a few military personnel", so it's retelling of rehearsal, even more trust.
3. "government plan was to mix the meat from Chernobyl-area cattle with the uncontaminated meat", that's seems plausible. contamination of meat was "up to 1,0*10-6 Ci/kg" (according to widely cited secret "annex 10 of protocol #32"), it's 300 times of banana equivalent. from the same scary stories fansites, there were 34 Кtons of bad meat (it's one thousand of railway carriages btw). so not so dramatic, definitely not a near nuclear blast.
4. [computer was] "located in a building close to the railroad tracks". inverse-square law, walls, exposure time from a moving train etc. i would more believe in vibration caused by trains as a cause of problems.
5. the SM-1800 computer: KR580VM80A cpu (8080 clone) is 6 mkm and 2 MHz (compare to modern 10 nm and 2 GHz), K565RU3A ram (4116 analog, like in ZX Spectrum), and other components. even the computer case was solid metal. i can not tell you precisely, but there should be a powerful source of radiation to crash such computer reproducibly.
so no, i don't buy it. sorry for hurting your radio- and russo-fobias.
That's a mighty big chip you have on your shoulder there. Sorry to disappoint your ego but I was curious as to your objections. They're... not very good.
The inverse square law one is the worst - you make tons of assumptions about the building and then screw up the basic science. A few inches of lead is enough to block gamma rays -- which is why you wear a lead apron when you get x-rays -- but at least 3 meters of concrete are needed to stop them.
Assuming a standard brick or breeze block building and normal computer metal case plus a meter or two of separation (common at railways in Europe at least, when stuff is going by slowly) that wouldn't stop much gamma radiation at all.
Other than "it's a semi-anonymous story" you really don't have much in the way of objection here.