Author

Topic: Block in future timestamp ? (Read 1235 times)

sr. member
Activity: 362
Merit: 262
September 03, 2015, 05:03:29 AM
#10
Yes.  It happens all the time.

Pardon the pun?
hero member
Activity: 692
Merit: 569
August 30, 2015, 05:57:23 AM
#9
Quote
The only thing the timestamp is intended to be used for is calculating the new proof-of-work difficulty every 2016 blocks.  A variation in timestamps of 120 minutes or so isn't a problem for that use case, and it means that the decentralized peer-to-peer system doesn't need a centralized source of time that everyone can agree on.

Thanks. Make sense !
legendary
Activity: 3472
Merit: 4801
August 30, 2015, 03:01:25 AM
#8
What is rational behind this design,

The only thing the timestamp is intended to be used for is calculating the new proof-of-work difficulty every 2016 blocks.  A variation in timestamps of 120 minutes or so isn't a problem for that use case, and it means that the decentralized peer-to-peer system doesn't need a centralized source of time that everyone can agree on.

This could lead to cases where timestamp of block n < timestamp of block n-1

Yes.  It happens all the time.
legendary
Activity: 1260
Merit: 1019
August 30, 2015, 02:50:57 AM
#7
bc.i has a bug

this site always shows the same timestamp for Block Timestamp and Received Time
hero member
Activity: 692
Merit: 569
August 29, 2015, 01:06:46 AM
#6
Timestamps are unreliable in a peer to peer environment. The only true 'time stamping' is POW.

So nodes will accept blocks with timestamp > current time ?

There is a wide bounds for acceptability, I can't remember the exact bounding they use.
I think the bounds are 120 minutes either way of the node's time.

What is rational behind this design, I understand 1 min or so , for time lags on node . If block timestamp is so unreliable, what is the point of having it in the block header. This could lead to cases where timestamp of block n < timestamp of block n-1
staff
Activity: 3458
Merit: 6793
Just writing some code
August 28, 2015, 09:39:18 AM
#5
Timestamps are unreliable in a peer to peer environment. The only true 'time stamping' is POW.

So nodes will accept blocks with timestamp > current time ?

There is a wide bounds for acceptability, I can't remember the exact bounding they use.
I think the bounds are 120 minutes either way of the node's time.
legendary
Activity: 1008
Merit: 1007
August 28, 2015, 08:48:37 AM
#4
Timestamps are unreliable in a peer to peer environment. The only true 'time stamping' is POW.

So nodes will accept blocks with timestamp > current time ?

There is a wide bounds for acceptability, I can't remember the exact bounding they use.
hero member
Activity: 692
Merit: 569
August 28, 2015, 08:13:59 AM
#3
Timestamps are unreliable in a peer to peer environment. The only true 'time stamping' is POW.

So nodes will accept blocks with timestamp > current time ?
legendary
Activity: 1008
Merit: 1007
August 28, 2015, 07:34:06 AM
#2
Timestamps are unreliable in a peer to peer environment. The only true 'time stamping' is POW.
hero member
Activity: 692
Merit: 569
August 28, 2015, 07:31:14 AM
#1
Today I found a block having timestamp which is > current time . Found this on my bitcoin node

Just to be sure I googled this on blockchain.info and found similar result:



Blockr had similar timestamp in the future ... but tradeblock was showing a past timestamp as expected

Wondering if somone could throw some light on this, is this expected ?
Jump to: