| Science and Tech Discuss questions of science, science and progress, that do not speak as loud as your heart. Also, tech stuff. |
![]() |
| | LinkBack | Thread Tools | Display Modes |
| | #1 (permalink) |
| Senior Member ![]() ![]() ![]()
Mayan Time Lord
Join Date: Apr 2008 Location: Cali
Posts: 1,992
My Mood: SL Join Date: 10/12/2006
Business: [H]arsh Styles
Client: Always changing, and too lazy to edit. | An Extra second will be added at Midnight Saturday. Be sure to adjust your Atomic Clocks! ![]() Don't get left in the past... The Science Behind Saturday's Leap Second - Yahoo! News On a more serious note, it seems there's some actual controversy around this: http://www.madore.org/~david/compute...p-seconds.html I wonder what LL will do... |
| | |
| |
| | #2 (permalink) |
| - ![]() ![]() ![]() ![]() ![]() Join Date: Jun 2008 Location: Darkly Cute :: Ferguson
Posts: 6,916
My Mood: SL Join Date: November, 2007
Business: Darkly Cute Client: NOT the same one I used yesterday... (<.<) | GPS clocks don't recognize the silly little leap second. =^-^=
__________________ Link |
| | |
| | #3 (permalink) | ||
| Senior Member ![]() ![]() ![]()
Mayan Time Lord
Join Date: Apr 2008 Location: Cali
Posts: 1,992
My Mood: SL Join Date: 10/12/2006
Business: [H]arsh Styles
Client: Always changing, and too lazy to edit. | I was reading some Google forums about this and it seems their Site Reliability Engineering Group (SREG) decided lying about the time was best: Quote:
Quote:
| ||
| | |
| | #5 (permalink) | |||
| That Bitch ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() *SLU Supporter* ![]() ![]() ![]() ![]() ![]()
Innocent as far as you know
Join Date: Nov 2011 Location: Online
Posts: 6,188
My Mood: SL Join Date: late 04... that account is deleted now | Quote:
__________________ - These eyes can do more than see Quote:
Quote:
| |||
| | |
| 1 User Agreed: |
| | #6 (permalink) |
| ┌( ಠ_ಠ)┘ ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() *SLU Supporter* ![]() ![]()
animating
| wot???? i find this thread more fun Be Sure to Adjust your Cocks Saturday! |
| | |
| 1 User Said Yay!: |
| 2 Users Laughed: |
| | #7 (permalink) | |
| Senior Member ![]() ![]() ![]()
Mayan Time Lord
Join Date: Apr 2008 Location: Cali
Posts: 1,992
My Mood: SL Join Date: 10/12/2006
Business: [H]arsh Styles
Client: Always changing, and too lazy to edit. | Quote:
Official Google Blog: Time, technology and leaping seconds When you consider the volume of traffic Google handles, having a second repeat twice could potentially cause issues. I think they just want to be more safe than sorry from what I read. | |
| | |
| | #8 (permalink) |
| That Bitch ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() *SLU Supporter* ![]() ![]() ![]() ![]() ![]()
Innocent as far as you know
Join Date: Nov 2011 Location: Online
Posts: 6,188
My Mood: SL Join Date: late 04... that account is deleted now | just seems very wrong to rely on that much precision against the real world clock... NTP is supposed to be more robust than that and it's ZOMG SHIT IS SERIOUSLY BORKED! threshold something like 15 minutes.... |
| | |
| | #9 (permalink) |
| Senior Member ![]() ![]()
This space for rent
| But but but, adding a leap second doesn't require repeating a second.... When the leap second is added the minute in question counts from 0 to 60 instead of 0 to 59 (at least, that's how I recall the NTP standard specifying it). That said, I can see where a lot of software systems are going to go SPUNG when they query the time and get 11:59:60 for the time... Smearing the leap second over a few seconds seems like a clever approach though it would be good to have that as part of the official standard I think, that way any systems that do rely on NTP to maintain sub-second synchronization can depend on it even during a leap second. |
| | |
| | #10 (permalink) |
| Senior Member ![]() ![]() ![]()
Mayan Time Lord
Join Date: Apr 2008 Location: Cali
Posts: 1,992
My Mood: SL Join Date: 10/12/2006
Business: [H]arsh Styles
Client: Always changing, and too lazy to edit. | Seems many experienced Linux Crashes due to the Leap Second. Never underestimate the ability for a system to fail! ![]() debian - Anyone else experiencing high rates of Linux server crashes during a leap second day? - Server Fault |
| | |
| | #11 (permalink) |
| That Bitch ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() *SLU Supporter* ![]() ![]() ![]() ![]() ![]()
Innocent as far as you know
Join Date: Nov 2011 Location: Online
Posts: 6,188
My Mood: SL Join Date: late 04... that account is deleted now | :: laughs :: if you read down there's a perfect description of what caused the problem.... the source of the problem was.... trying to solve the problem! =D the local machine, rather than try to update it's clock directly from the NTP tries to actually handle the leap second on it's own, when it's been flagged. to do so requires extra code and handling.... and that code had a potential race condition that can cause a freeze-crash in the OS. the thing is... that extra code for handling leap seconds is part of the NTP spec.... you are supposed to do that.... and yet, if the extra code wasn't there on the local side, you'd simply synch up as normal when NTP had a time that's 1 second off. so it turned out that an overthought original solution led to a worse problem than it was supposed to solve. ETA: and that dependency I was wondering about... turns out to be NTP itself =D |
| | |
| 1 User Laughed: |
| | #12 (permalink) | |
| feral ballerina ![]() ![]()
Hello, I'm the Groove
operator. Please hold the
line.
Join Date: May 2009 Location: Left coast
Posts: 857
My Mood: SL Join Date: March 2004
Business: D3volution
| Quote:
| |
| | |
| | #16 (permalink) |
| feral ballerina ![]() ![]()
Hello, I'm the Groove
operator. Please hold the
line.
Join Date: May 2009 Location: Left coast
Posts: 857
My Mood: SL Join Date: March 2004
Business: D3volution
| One second sooner than you thought it would, too. That's gonna be a big surprise, come next June. |
| | |
| 2 Users Laughed: |
![]() |
| Tags |
| leap second |
| Thread Tools | |
| Display Modes | |
| |
| |