To: bitcoin-dev@lists.linuxfoundation.org
References: <CABsx9T1E1s=4h-SxLTOAXK4GniZrUekcEb6zDdTDFG+h7X98MA@mail.gmail.com>
	<1438640036.2828.0.camel@auspira.com>
	<CABsx9T2A-Mz9z=TTifbL2_sKCDvy8coRpNse+0vff6EbXbp8cg@mail.gmail.com>
	<BF420F3B-044C-46F6-8880-FEEB9A3DC748@gmx.com>
	<3162BC78-EC0B-4DAA-A472-D143389DDD8A@hashingit.com>
	<7F9283D5-5F0F-4318-BE64-A1C20A5B7606@gmx.com>
	<2312E340-EA7D-48DC-B3FF-319D6AF9E955@hashingit.com>
From: Tom Harding <tomh@thinlink.com>
Message-ID: <55C2A012.7080908@thinlink.com>
Date: Wed, 5 Aug 2015 16:45:22 -0700
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:38.0) Gecko/20100101
	Thunderbird/38.1.0
MIME-Version: 1.0
In-Reply-To: <2312E340-EA7D-48DC-B3FF-319D6AF9E955@hashingit.com>
Content-Type: text/plain; charset=windows-1252; format=flowed
Content-Transfer-Encoding: 7bit
Subject: Re: [bitcoin-dev] "A Transaction Fee Market Exists Without a Block
 Size Limit"--new research paper suggests
Precedence: list

On 8/5/2015 3:44 PM, Dave Hudson via bitcoin-dev wrote:
> I do suspect that if we were to model this more accurately we might be 
> able to infer the "typical" propagation characteristics by measuring 
> the deviation from the expected distribution. 

The paper models propagation using a single time value that is a 
function of block size.  Modeling the propagation distribution (which is 
totally separate from the poisson model of block production) would add a 
lot of complexity and my guess is the outcome would be little changed.


>> On 5 Aug 2015, at 15:15, Peter R <peter_r@gmx.com> wrote:
>> Although a miner may not orphan his own block, by building on his own block he may now orphan two blocks in a row.  At some point, his solution or solutions must be communicated to his peers.

Why complicate the analysis by assuming that a miner who finds two 
blocks sequentially does not publish the first, or that other miners 
would orphan miner's first block unless both were very quick?  In 
general you don't consider anything beyond 1 block in the future, which 
seems fine.


>>> I suspect this may well change some of the conclusions as larger block makers will definitely be able to create larger blocks than their smaller counterparts.
>> It will be interesting to see.  I suspect that the main result that "a healthy fee market exists" will still hold (assuming of course that a single miner with >50% of the hash power isn't acting maliciously).  Whether miners with larger value of h/H have a profit advantage, I'm not sure (but that was outside the scope of the paper anyways).

Correcting for non-orphaning of one's own blocks could be as simple as 
adding a factor (1 - h/H) to equation 4, which it appears would leave 
hashpower as an independent variable in the results.  But at worst, the 
discussion can be considered to apply directly only to low-hashpower 
miners right now.

Overall, the paper does not predict big changes to per/kb fees or spam 
costs for the kinds of block sizes being discussed for the immediate 
future (8MB).  But it does conclude that these fees will rise, not fall, 
with bigger blocks.

Also it is welcome that this paper actually mentions the bitcoin 
exchange rate as a factor in relation to block size (it points out that 
a spam attack is much more expensive in fiat terms today than it was 
years ago).