01:35:07ghtdak:ghtdak has left #bitcoin-wizards
03:08:15jgarzik_:jgarzik_ is now known as jgarzik
09:38:10wallet421:wallet421 is now known as wallet42
09:44:18nikitab_:nikitab_ is now known as nikitab
10:23:45fanquake:fanquake has left #bitcoin-wizards
10:43:33wallet42:wallet42 is now known as Guest6086
10:43:33wallet421:wallet421 is now known as wallet42
11:06:05[Tristan]:[Tristan] is now known as Guest12331
11:49:03Guest12331:Guest12331 is now known as Tristna
11:49:30Tristna:Tristna is now known as _Tristan_
12:22:55_Tristan_:_Tristan_ is now known as [Tristan]
14:42:42DoctorBTC_:DoctorBTC_ is now known as DoctorBTC
16:24:02maaku:maaku is now known as Guest11836
16:25:07Guest11836:Guest11836 has left #bitcoin-wizards
16:25:49Guest11836:Guest11836 is now known as maaku
16:53:43maaku:maaku is now known as Guest81467
16:56:40Guest81467:Guest81467 is now known as maaku
17:02:06dgenr8:with that analogy in mind, i tried, and failed miserably, to come up with a way of ordering transactions using collision detection. so i realized why we're not all on one big LAN. heh.
17:04:07dsnrk:that is at the core what mining provides, ordering
17:04:29sipa:if it were possible to consistently order events generated in different geographic locations, we wouldn't need a blockchain :)
17:04:52sipa:everyone would just only accept the first non-conflicting one
17:05:18dgenr8:the ordering that comes from the blockchain is quite chunky
17:06:18sipa:use a higher frequency of blocks if you need higher resolution, but at the cost of higher reorganizationnchance
17:06:19dgenr8:i think it's necessary to focus directly on the goals that a theoretical distributed clock would achieve, and find ways to move towards them, given actual network performance and no distributed clock
17:06:36sipa:that's not possible
17:06:50sipa:it's always a balance between certainty and resolution
17:06:52dgenr8:not with 100% certainty
17:07:30sipa:the closer your resilution gets to the network diameter, the more uncertain the value of a block is
17:08:07sipa:if you want a distributed clock, use a central timestamper
17:08:23dgenr8:i'm talking about improving bitcoin here
17:08:26sipa:(who gets to censor, and is a central point of failure)
17:08:37sipa:well, this channel is not about bitcoin :)
17:08:57sipa:or at least not near-future developments
17:09:35dgenr8:a key fact is that the network will always perform better than mining activity as an abstract process. because mining activity is built on the network
17:09:57sipa:higher accuracy, but it lacks consensus
17:10:00dgenr8:so shorter blocktimes may not be the solution to improving resolution
17:10:43sipa:and it is not necessary that mining takes its transactions from the network
17:10:58sipa:that's how the current flood broadcasting of transactions is intended
17:11:23sipa:but it could just be that transactions are sent to, and paid for, privately to miners
17:12:03sipa:(not saying that is a good thing, just clarifying that transaction broadcasting is not inseparably linked to the blockchain as a system)
17:12:03dgenr8:hmm. then why did satoshi include it
17:15:50sipa:satoshi assumed that every full node would be a miner :)
17:28:49jgarzik:Some tweets RE mining that you guys might be interested in, up at https://twitter.com/jgarzik
17:37:28dgenr8:jgarzik: YES, strongly agreed regarding mining to influence policy. The foundation could even do this to ensure smooth acceptance of a soft fork.
20:37:11andrew_:andrew_ is now known as justanotheruser
21:50:22helo_:helo_ is now known as helo
22:57:30stqism:stqism is now known as NikolaiToryzin
23:00:14NikolaiToryzin:NikolaiToryzin is now known as stqism
23:02:27stqism:stqism is now known as irunbsd
23:12:42irunbsd:irunbsd is now known as stqism