From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from sog-mx-2.v43.ch3.sourceforge.com ([172.29.43.192] helo=mx.sourceforge.net) by sfs-ml-4.v29.ch3.sourceforge.com with esmtp (Exim 4.76) (envelope-from ) id 1WoMhw-0003KY-KG for bitcoin-development@lists.sourceforge.net; Sun, 25 May 2014 00:54:36 +0000 X-ACL-Warn: Received: from fallback2.mail.ox.ac.uk ([129.67.1.167]) by sog-mx-2.v43.ch3.sourceforge.com with esmtp (Exim 4.76) id 1WoMhu-0008OI-P2 for bitcoin-development@lists.sourceforge.net; Sun, 25 May 2014 00:54:36 +0000 Received: from relay12.mail.ox.ac.uk ([129.67.1.163]) by fallback2.mail.ox.ac.uk with esmtp (Exim 4.80) (envelope-from ) id 1WoLow-000237-7m for bitcoin-development@lists.sourceforge.net; Sun, 25 May 2014 00:57:46 +0100 Received: from hub01.nexus.ox.ac.uk ([163.1.154.218] helo=HUB01.ad.oak.ox.ac.uk) by relay12.mail.ox.ac.uk with esmtp (Exim 4.80) (envelope-from ) id 1WoLoq-0004x2-eo for bitcoin-development@lists.sourceforge.net; Sun, 25 May 2014 00:57:40 +0100 Received: from MBX03.ad.oak.ox.ac.uk ([169.254.3.44]) by HUB01.ad.oak.ox.ac.uk ([163.1.154.92]) with mapi id 14.03.0169.001; Sun, 25 May 2014 00:57:40 +0100 From: Jonathan Levin To: "bitcoin-development@lists.sourceforge.net" Thread-Topic: Cut-through propagation of blocks Thread-Index: AQHPd6vxYIXrqyO45kqgPDYZ/wCKiw== Date: Sat, 24 May 2014 23:57:39 +0000 Message-ID: <86C2A90A-6619-45CD-94EC-009456471050@sant.ox.ac.uk> References: In-Reply-To: Accept-Language: en-GB, en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [172.16.150.237] Content-Type: multipart/alternative; boundary="_000_86C2A90A661945CD94EC009456471050santoxacuk_" MIME-Version: 1.0 X-Spam-Score: 0.3 (/) X-Spam-Report: Spam Filtering performed by mx.sourceforge.net. See http://spamassassin.org/tag/ for more details. -0.7 RP_MATCHES_RCVD Envelope sender domain matches handover relay domain 1.0 HTML_MESSAGE BODY: HTML included in message X-Headers-End: 1WoMhu-0008OI-P2 Subject: Re: [Bitcoin-development] Cut-through propagation of blocks X-BeenThere: bitcoin-development@lists.sourceforge.net X-Mailman-Version: 2.1.9 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 25 May 2014 00:54:36 -0000 --_000_86C2A90A661945CD94EC009456471050santoxacuk_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable I have done some work on incentives arising from block propagation times an= d it turns out that Bitcoin is already quite good at establishing the prima= cy of blocks by time despite what people think. Part of the reason for this= is the way that partitions on the network evolve as a block is propagated.= Typically at the moment, blocks reach over 50% of the network in 5 seconds= . Reach being defined as a node receiving and validating a block. If we mak= e an assumption that the hashing power of the network is uniformly distribu= ted over the nodes (I know it is not a good assumption but can discuss it o= ff the list). Then 50% of the hashing power are already building a block th= at builds on top of the block that is already circulating. The probability = that there is a collision on the network therefore falls fast and then the = probability that the miner who propagated the first block wins given a coll= ision occurs is rising. I think that block propagation times might actually= be a bigger issue for miners who are less well connected to the network in= the sense that they spend more time mining redundant problems and during t= hat time may find blocks to compete with blocks that are already spreading = throughout the network. I have a paper that models this more formally and has some numerical simula= tions but cannot publish it on the internet at present (University Regulati= ons) but I am happy to share a version privately if anyone is interested. Best, Jonathan -- Jonathan Levin Co-Founder Coinometrics http://www.coinometrics.com/ Postgraduate Economist | St Antony's College | Oxford University @jony_levin @Coinometrics --_000_86C2A90A661945CD94EC009456471050santoxacuk_ Content-Type: text/html; charset="us-ascii" Content-ID: <19CC0F2BD04F494CAFEF4C6E23F7224A@ad.oak.ox.ac.uk> Content-Transfer-Encoding: quoted-printable I have done some work on incentives arising from block propagation times an= d it turns out that Bitcoin is already quite good at establishing the prima= cy of blocks by time despite what people think. Part of the reason for this= is the way that partitions on the network evolve as a block is propagated. Typically at the moment, blocks r= each over 50% of the network in 5 seconds. Reach being defined as a node re= ceiving and validating a block. If we make an assumption that the hashing p= ower of the network is uniformly distributed over the nodes (I know it is not a good assumption but can dis= cuss it off the list). Then 50% of the hashing power are already building a= block that builds on top of the block that is already circulating. The pro= bability that there is a collision on the network therefore falls fast and then the probability that the mine= r who propagated the first block wins given a collision occurs is rising. I= think that block propagation times might actually be a bigger issue for mi= ners who are less well connected to the network in the sense that they spend more time mining redundant pro= blems and during that time may find blocks to compete with blocks that are = already spreading throughout the network. 

I have a paper that models this more formally and has some numerical s= imulations but cannot publish it on the internet at present (University Reg= ulations) but I am happy to share a version privately if anyone is interest= ed.

Best,

Jonathan

--
Jonathan Levin
Co-Founder Coinometrics
http://www.coinometrics.com/
Postgraduate Economist | St Antony's College | Oxford University
@jony_levin
@Coinometrics