Received-SPF: pass (sog-mx-2.v43.ch3.sourceforge.com: domain of yahoo.com
	designates 67.195.87.187 as permitted sender)
	client-ip=67.195.87.187; envelope-from=rdwnj@yahoo.com;
	helo=nm47-vm9.bullet.mail.gq1.yahoo.com; 
Date: Mon, 25 May 2015 20:14:17 +0000 (UTC)
From: Ron <rdwnj@yahoo.com>
To: "bitcoin-development@lists.sourceforge.net"
	<bitcoin-development@lists.sourceforge.net>
Message-ID: <542308084.933798.1432584857775.JavaMail.yahoo@mail.yahoo.com>
MIME-Version: 1.0
Content-Type: multipart/alternative; 
	boundary="----=_Part_933797_1204143679.1432584857773"
Subject: [Bitcoin-development] alternatives to the 20MB block limit,
	measure first!
Precedence: list
Reply-To: Ron <rdwnj@yahoo.com>

------=_Part_933797_1204143679.1432584857773
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Hello all,

With all the discussion about the Block size limit, I thought it would be i=
nteresting to measure, in some sense, the average Tx size.=C2=A0 Then given=
 a fixed average block period (Bp) of 10 minutes (i.e 600 seconds), all one=
 needs to do to estimate an average block size is ask the question: what av=
erage transaction rate (tps) do you want?

So for tps ~ 10 (Tx/sec) and an average transaction size (avgTxSz) of 612 B=
ytes (last ten blocks up to block 357998 (2:05pm EDT 5/25/2015) we have a b=
lock size of 612 * 10 * 600 =3D 3,672,000 Bytes

Alternatively, given an avgTxSz ~612 and maxBl =3D 1,000,000 we have (maxBl=
 / avgBlSz) / Bp is the actual current max tps, which is ~2.72 tps.

The avgBlSz for the 10 blocks up to block # 357999 is ~ 576 Bytes, so the c=
urrent possible tps is ~2.89 and the maxBL for a tps =3D 10 is 3,456,000 by=
tes.

So I think one should state one's assumed tps and a measured or presumed av=
gTxSz before saying what a maxBl should be. So for a maxBl ~20,000,000 Byte=
s and a current avgTxSz ~600 Bytes, the tps ~55.5 FWIW

Ron (aka old c coder)


------=_Part_933797_1204143679.1432584857773
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<html><body><div style=3D"color:#000; background-color:#fff; font-family:ti=
mes new roman, new york, times, serif;font-size:14px"><div dir=3D"ltr" id=
=3D"yui_3_16_0_1_1432584468852_2493">Hello all,<br style=3D"" class=3D""><b=
r style=3D"" class=3D"">With all the discussion about the Block size limit,=
 I thought it would be interesting to measure, in some sense, the average T=
x size.&nbsp; Then given a fixed average block period (Bp) of 10 minutes (i=
.e 600 seconds), all one needs to do to estimate an average block size is a=
sk the question: what average transaction rate (tps) do you want?<br style=
=3D"" class=3D""><br style=3D"" class=3D"">So for tps ~ 10 (Tx/sec) and an =
average transaction size (avgTxSz) of 612 Bytes (last ten blocks up to bloc=
k 357998 (2:05pm EDT 5/25/2015) we have a block size of 612 * 10 * 600 =3D =
3,672,000 Bytes<br style=3D"" class=3D""><br style=3D"" class=3D"">Alternat=
ively, given an avgTxSz ~612 and maxBl =3D 1,000,000 we have (maxBl / avgBl=
Sz) / Bp is the actual current max tps, which is ~2.72 tps.<br style=3D"" c=
lass=3D""><br style=3D"" class=3D"">The avgBlSz for the 10 blocks up to blo=
ck # 357999 is ~ 576 Bytes, so the current possible tps is ~2.89 and the ma=
xBL for a tps =3D 10 is 3,456,000 bytes.<br style=3D"" class=3D""><br style=
=3D"" class=3D"">So I think one should state one's assumed tps and a measur=
ed or presumed avgTxSz before saying what a maxBl should be. So for a maxBl=
 ~20,000,000 Bytes and a current avgTxSz ~600 Bytes, the tps ~55.5 FWIW<br =
style=3D"" class=3D""><br style=3D"" class=3D"">Ron (aka old c coder)<br></=
div><div style=3D"" class=3D"" id=3D"yui_3_16_0_1_1432584468852_2493"><br s=
tyle=3D"" class=3D""></div></div></body></html>
------=_Part_933797_1204143679.1432584857773--