|
Title: Can Bitcoin be affected by an extra second? Post by: nachoig on February 16, 2015, 03:28:22 AM Explaining: because of the Earth's rotation speed, 2015 will have an extra second. According to this article, this extra second can lead to some instabilities over the internet. So, for example, can this affect the interval between Bitcoin blocks?
http://www.theverge.com/2015/1/7/7508651/leap-second-2015-earths-rotation-slowing Title: Re: Can Bitcoin be affected by an extra second? Post by: DeathAndTaxes on February 16, 2015, 03:30:39 AM No. Bitcoin already has a very loose time synchronization. Timestamps can be off as much as two hours and still be considered valid. An extra second isn't even detectable.
Title: Re: Can Bitcoin be affected by an extra second? Post by: Foxpup on February 16, 2015, 04:25:13 AM To elaborate, systems that assume Unix timestamps are always sequential will be mistaken about the ordering of events that take place during a leap second and the following second, since the two seconds are indistinguishable from the timestamps. For example, an event which occurs at 11:59:60.8 PM will have a later timestamp than an event that takes places half a second later at 12:00:00.3 AM, since 11:59:60 PM and 12:00:00 AM are the same in Unix time. (And some systems, when dealing with human-readable time formats, will give an error on seeing the 60.8 seconds, since such a timestamp is "obviously" invalid.)
Bitcoin has none of these problems. It does not use timestamps to determine the ordering of events, with out-of-order timestamps being perfectly valid, and a single second will not make any appreciable difference to the block interval. Title: Re: Can Bitcoin be affected by an extra second? Post by: Beymond on February 16, 2015, 05:07:30 AM First i was like why is there even a question like this
After reading this Exactly :D To elaborate, systems that assume Unix timestamps are always sequential will be mistaken about the ordering of events that take place during a leap second and the following second, since the two seconds are indistinguishable from the timestamps. For example, an event which occurs at 11:59:60.8 PM will have a later timestamp than an event that takes places half a second later at 12:00:00.3 AM, since 11:59:60 PM and 12:00:00 AM are the same in Unix time. (And some systems, when dealing with human-readable time formats, will give an error on seeing the 60.8 seconds, since such a timestamp is "obviously" invalid.) Bitcoin has none of these problems. It does not use timestamps to determine the ordering of events, with out-of-order timestamps being perfectly valid, and a single second will not make any appreciable difference to the block interval. Title: Re: Can Bitcoin be affected by an extra second? Post by: samson on February 16, 2015, 12:27:00 PM There's a whole mailing list dedicated to leap seconds and how they should be handled / standardised.
Title: Re: Can Bitcoin be affected by an extra second? Post by: nachoig on February 16, 2015, 08:57:33 PM Thanks for the explanation people. :)
|