Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The code worked differently when the moon was full

hanselman.com

Post reply on HN
Print view View original

51–60 of 175 posts

Re: The code worked differently when the moon was full

#51

Post by lmilcin »

I have seen interference from one part of set top box to cause noise on flash input lines and sometimes issue a command to clear flash on the device.

Months of debugging, dozen people involved, tens of thousands of devices bricked, tens of millions lost.

All due to a single line of code that configured flash to not require special magic before each command. This feature made to improve resistance to interference also hindered performance. Somebody thought it a good idea to disable to get some points for improved performance.

Re: The code worked differently when the moon was full

#53

Post by HelixEndeavor »

vidanay wrote: post #26

We once had a customer who would call us in a panic a couple of times a year saying our inspection equipment was experiencing unusually high false rejects and they were generating very high scrap rate. By the time we got a technician on site the next day, everything was working flawlessly and the customer couldn't reproduce the problem either. This went on for almost three years with various levels of escalation to t…

Indiana Jones discovering the location of the Ark.

Re: The code worked differently when the moon was full

#54

Post by tclancy »

chongli wrote: post #49

I'm immediately reminded of NetHack [1] (which you can play online here [2])! Real-life phase of the moon has a small but important effect on gameplay in NetHack. The public game server on alt.org even tracks the phase of the moon to provide a handy reference. A little bit disappointing to discover that the code from the article does not actually depend on the phase of the moon. I'm really interested to see the other…

Slightly related, I once wrote code for the old Kill Screen site that subtly altered the graphics of a review of a game themed around lunar stuff based on the current phase of the moon.

God bless gusts of random math people leave about.

Re: The code worked differently when the moon was full

#55

Post by walrus01 »

mhandley wrote: post #12

The phase of the moon really can affect performance. A friend of mine worked on wireless links in Scotland and was struggling with loss at certain times of day, but not exactly the same time every day. When they graphed loss against time, the pattern was really periodic over many days. The periodicity turned out to be 12 hours 25 minutes, which they eventually realized is exactly the time between low tides. The probl…

point to point microwave link sounds like the bottom part of the fresnel zone was scraping the water - bad engineering design from the outset, tides or no tides. Not a good idea to do unless you have absolutely no economical way of getting one or both ends of the link higher.

Re: The code worked differently when the moon was full

#56

Post by walrus01 »

ColinWright wrote: post #36
mhandley wrote: post #12

The phase of the moon really can affect performance. A friend of mine worked on wireless links in Scotland and was struggling with loss at certain times of day, but not exactly the same time every day. When they graphed loss against time, the pattern was really periodic over many days. The periodicity turned out to be 12 hours 25 minutes, which they eventually realized is exactly the time between low tides. The probl…

For those who are interested, this is why you usually have two dishes/aerials, vertically displaced, so that when one has destructive interference between the direct and reflected signals, the other has constructive interference. I learned something about this when writing data compression and encryption software for radar surveillance systems, where there were multiple radars over a moderate coverage area, all sendi…

in modern point to microwave systems, unless the budget is really high, or the path is very long, it's rather unusual to have a vertical spatial diversity setup. much of the problem of losing link due to fade is accommodated by modern radios that have advanced variable modulation and FEC, which can operate anywhere between 4096QAM 5/6 and QPSK 1/2.

Re: The code worked differently when the moon was full

#57

Post by tabtab »

vidanay wrote: post #26

We once had a customer who would call us in a panic a couple of times a year saying our inspection equipment was experiencing unusually high false rejects and they were generating very high scrap rate. By the time we got a technician on site the next day, everything was working flawlessly and the customer couldn't reproduce the problem either. This went on for almost three years with various levels of escalation to t…

Our garage door opener has this problem. At very specific times of the year the sun confuses the infrared blockage sensor. The cause occurred to me when I lined up my eye to see what the sensor was seeing when it was failing, and I noticed the morning sun right next to the other end of the sensor stream. I moved a trash can to shade the sensor and it worked fine.

Re: The code worked differently when the moon was full

#58

Post by walrus01 »

m4rtink wrote: post #31
mhandley wrote: post #12

The phase of the moon really can affect performance. A friend of mine worked on wireless links in Scotland and was struggling with loss at certain times of day, but not exactly the same time every day. When they graphed loss against time, the pattern was really periodic over many days. The periodicity turned out to be 12 hours 25 minutes, which they eventually realized is exactly the time between low tides. The probl…

I heard a story about an astronomer loosing the chance to be the first to report a commet one cold winter night - just as he wanted to send the email to report it, the Internet connection was dead! He ran from the observatory to the nearest place with Internet connectivity, but by the time he sent the email from there, there was already a report from another astronomer elsewhere, a few minutes ago. Reason for the mys…

one of the reasons why free space optics gigabit and 10GbE ethernet laser based links are a tiny niche product in the ISP space, millimeter wave (71-86 GHz) fills the purpose much better.

also they don't do as well in rain as something that can adapt modulation.

Re: The code worked differently when the moon was full

#59

Post by walrus01 »

Socketier wrote: post #37
a-dub wrote: post #30

Earlier quoted context omitted.

i once lived in an apartment in an old building where if you turned on the light in the bathroom the dsl would lose sync. why this would occur i'll never know. (probably old telephone wiring wrapped around old 110v wiring? maybe? or who knows what kind of weird leakage/ground loops may have existed)

This happened to me as well, only with DSL. The strip lights on the bathroom mirror (newly installed) would disconnect the internet when turned on. Such a weird thing to troubleshoot when you have a few people living in the same house.

cheap AC to DC power supplies are very 'noisy' in the RF spectrum.
Post reply on HN