Author

Topic: Why is the block time 10 minutes? (Read 1582 times)

legendary
Activity: 1204
Merit: 1015
April 16, 2013, 07:21:04 PM
#17
If you wait for 6 confirmations on a 10 min block chain then with equivalent hashpower you should wait 24 blocks on a 2.5 min block chain.  If you are willing to accept 4 confirmations on a 2.5 minute blockchain then 1 confirmation on a 10 minute blockchain provides equivalent security.  
This was debunked some time ago. Unfortunately, I don't have a link, but someone made a paper about it. Basically, the more blocks that are produced, regardless of how long it takes, the less volatility there is in the apparent hash power of the network. In short, someone with 1% of the hash power could quite easily create 1 block faster than the rest of the network, and maybe even 2 if they're lucky. 6 is much harder.
So 1 confirmation is as secure as 1 confirmation, no matter how fast the block time?
Yes, but only if you exclude the increased orphan rate that would result from faster block times. Hence why 1-second blocks have proven to be a bad idea (many alt-chains started with very low difficulty and didn't start stabilizing until the difficulty jumped up to take 3 seconds per block).

This was debunked some time ago. Unfortunately, I don't have a link, but someone made a paper about it.
Was this it? https://bitcoil.co.il/Doublespend.pdf
Indeed it was! Thanks!
newbie
Activity: 41
Merit: 0
April 16, 2013, 07:14:16 PM
#16
This was debunked some time ago. Unfortunately, I don't have a link, but someone made a paper about it.
Was this it? https://bitcoil.co.il/Doublespend.pdf
newbie
Activity: 27
Merit: 0
April 16, 2013, 07:01:10 PM
#15
it is picked arbitrarily and can be changed by a vote.
newbie
Activity: 41
Merit: 0
April 16, 2013, 06:51:23 PM
#14
If you wait for 6 confirmations on a 10 min block chain then with equivalent hashpower you should wait 24 blocks on a 2.5 min block chain.  If you are willing to accept 4 confirmations on a 2.5 minute blockchain then 1 confirmation on a 10 minute blockchain provides equivalent security.  
This was debunked some time ago. Unfortunately, I don't have a link, but someone made a paper about it. Basically, the more blocks that are produced, regardless of how long it takes, the less volatility there is in the apparent hash power of the network. In short, someone with 1% of the hash power could quite easily create 1 block faster than the rest of the network, and maybe even 2 if they're lucky. 6 is much harder.
So 1 confirmation is as secure as 1 confirmation, no matter how fast the block time?
legendary
Activity: 1204
Merit: 1015
April 16, 2013, 05:32:36 PM
#13
If you wait for 6 confirmations on a 10 min block chain then with equivalent hashpower you should wait 24 blocks on a 2.5 min block chain.  If you are willing to accept 4 confirmations on a 2.5 minute blockchain then 1 confirmation on a 10 minute blockchain provides equivalent security.  
This was debunked some time ago. Unfortunately, I don't have a link, but someone made a paper about it. Basically, the more blocks that are produced, regardless of how long it takes, the less volatility there is in the apparent hash power of the network. In short, someone with 1% of the hash power could quite easily create 1 block faster than the rest of the network, and maybe even 2 if they're lucky. 6 is much harder.
sr. member
Activity: 475
Merit: 255
April 14, 2013, 05:44:43 PM
#12
Newbie but related question:

If I start paying for my groceries with bitcoins, does the store security guy not let me leave with my milk for 10 minutes?  Wont my milk spoil?



Milk does not spoil in 10 minutes :-].

But to answer your question... It is up to grocery. Most probably they will be ok with 0-confirmation for milk. Perhaps security guy will wait for several seconds. You would have to invest quite an effort to double spend, effort unworthy of milk. And if the grocery detects double spend then they have your face on camera or security guy remembers you. And you will not be welcome in the grocery next time.
donator
Activity: 1218
Merit: 1080
Gerald Davis
April 14, 2013, 03:25:27 PM
#11
Newbie but related question:

If I start paying for my groceries with bitcoins, does the store security guy not let me leave with my milk for 10 minutes?  Wont my milk spoil?



If the grocery store makes you wait 48 hours before they receive funds from your credit card won't you be late for work?
phk
newbie
Activity: 28
Merit: 0
April 14, 2013, 03:18:28 PM
#10
Newbie but related question:

If I start paying for my groceries with bitcoins, does the store security guy not let me leave with my milk for 10 minutes?  Wont my milk spoil?

donator
Activity: 1218
Merit: 1080
Gerald Davis
April 14, 2013, 02:02:47 PM
#9
So, as I said before, it's completely arbitrary?

Not completely. The time has to be long enough for every node to verify the block. 10 minutes was chosen by Satoshi as a compromise between network security and confirmation times. In hindsight a smaller value might have been better, but not much smaller. With older versions of Bitcoin it sometimes took over a minute before every node received the block, and there needs to be a wide margin for safety.

This.  +1 rare for noob to get it that accurate.

To expand upon what Infinity said ...

Any block interval is a compromise.  10 min, 1 min, 60 min, etc.  There is no right or wrong.  It is a compromise.  Remember the actual block time will vary.  When blocks can't propagate the network fast enough and competing blocks are produced that results in orphans.  The % of orphans direct reduces the security of the network.  Currently Bitcoin w/ 10 minute blocks (and relatively small blocks) has about 1% orphan rate.  That means 1% of hashing power is wasted and doesn't improve security.  As blocks get larger the orphan rate will rise (although faster CPU and higher bandwidth connections improve the orphan rate).

10 minutes is a compromise between confirmation times and network security.  Really unless you are accepting 1 confirm txs faster block interval won't make tx validate faster.  If you wait for 6 confirmations on a 10 min block chain then with equivalent hashpower you should wait 24 blocks on a 2.5 min block chain.  If you are willing to accept 4 confirmations on a 2.5 minute blockchain then 1 confirmation on a 10 minute blockchain provides equivalent security.  

Really a shorter block interval only helps if your business accepts 1 confirm txs (because 1 confirm is always more secure than 0 confirms regardless of the block interval).
full member
Activity: 192
Merit: 100
April 14, 2013, 01:55:44 PM
#8
So, as I said before, it's completely arbitrary?

Not completely. The time has to be long enough for every node to verify the block. 10 minutes was chosen by Satoshi as a compromise between network security and confirmation times. In hindsight a smaller value might have been better, but not much smaller. With older versions of Bitcoin it sometimes took over a minute before every node received the block, and there needs to be a wide margin for safety.
You are totally correct, Sir.
newbie
Activity: 7
Merit: 0
April 14, 2013, 01:47:05 PM
#7
So, as I said before, it's completely arbitrary?

Not completely. The time has to be long enough for every node to verify the block. 10 minutes was chosen by Satoshi as a compromise between network security and confirmation times. In hindsight a smaller value might have been better, but not much smaller. With older versions of Bitcoin it sometimes took over a minute before every node received the block, and there needs to be a wide margin for safety.
newbie
Activity: 18
Merit: 0
April 14, 2013, 01:19:33 AM
#6
even though hash rates are going up aren't there more transactions per second in the blockchain to be hashed out meaning a zero sum increase or am I missing something in the scheme of it all ? Huh

Yes, you are. The amount of transactions in a block doesn't have an influence on how long it takes to find a hash that is below the current network threshold (difficulty level).

I thought what crypto canary was asking is why isn't the span five minutes?  What is "optimal" about having it be 10 minutes?
newbie
Activity: 23
Merit: 0
April 14, 2013, 12:51:23 AM
#5
even though hash rates are going up aren't there more transactions per second in the blockchain to be hashed out meaning a zero sum increase or am I missing something in the scheme of it all ? Huh

Yes, you are. The amount of transactions in a block doesn't have an influence on how long it takes to find a hash that is below the current network threshold (difficulty level).
full member
Activity: 134
Merit: 100
April 13, 2013, 11:12:21 PM
#4
even though hash rates are going up aren't there more transactions per second in the blockchain to be hashed out meaning a zero sum increase or am I missing something in the scheme of it all ? Huh
newbie
Activity: 41
Merit: 0
April 13, 2013, 09:52:02 PM
#3
The short answer it is a by product of how the network was configured from the get go.
So, as I said before, it's completely arbitrary?
mjc
hero member
Activity: 588
Merit: 500
Available on Kindle
April 13, 2013, 09:43:45 PM
#2
10 minutes is part of the pacing of the network.  A new block is found about every ten minutes.  The difficultly setting is there to keep the pace at 10 minutes.  This way there is a predictabl rate of new coins added into the system. 

For more information:  https://en.bitcoin.it/wiki/FAQ#Why_do_I_have_to_wait_10_minutes_before_I_can_spend_money_I_received.3F

That said, with the increasing network hash rate the frequency of blocks have been slightly more than 6 an hour, meaning faster confirmation times:
http://blockchain.info/charts/avg-confirmation-time

The short answer it is a by product of how the network was configured from the get go.
newbie
Activity: 41
Merit: 0
April 13, 2013, 09:35:55 PM
#1
One thing I haven't really been able to find out from the few weeks I've been interested in Bitcoin is why is there a 10 minute block time? That number seems completely arbitrary, and all it does is slow down transactions. The only argument I've read against shorter transaction times is that each conformation is less secure, but why does that matter, since it would take less time for subsequent conformations?
Jump to: