Hi John,
 
Let's examine a modest assumption:
 
2MB blocks and 10 seconds imply upto 100 tx/s at a bandwidth of 1.6 MBit/s and storage requirement of max 101GB/year. These should not create a problem even today. 
We could set minimum fee to 1/1000 of the block reward, close to Bitcoin's. ( Bitcoin's current average fee is 27000 satoshis, that is 1/771 of its 10 sec block reward (27000/(12.5/60*10^8)) )
 
With above parameters dilution would drop to 0% or less as soon as the network is fully utilized, since 1000 tx per block would burn at least the new reward in fees.
 
Current dilution would be a function of both earlier and current use of the network and oscillate around the 1/(years - 1) rate implied by the constant reward.
 
I really like that burning fees aligns the interest of those who want to use the network as store of value with those who want a high throughput transaction network.
 
Garrick Ollivander
 
Sent: Sunday, October 01, 2017 at 3:37 PM
From: "John Tromp" <john.tr...@gmail.com>
To: "Garrick Ollivander" <garrickollivan...@gmx.com>
Cc: "Seamus Finnigan" <seamus.finni...@protonmail.com>, "mimblewimble@lists.launchpad.net" <mimblewimble@lists.launchpad.net>
Subject: Re: [Mimblewimble] On block rewards
On Sun, Oct 1, 2017 at 3:10 PM, Garrick Ollivander
<garrickollivan...@gmx.com> wrote:

> Let's assume we want 0% dilution after grin reaches Visa like 2000 tx/s use
> use with a block time of 10 seconds,

That's completely unrealistic. Bandwidth wise, grin scales worse than bitcoin,
with transactions taking well over 2KB, mostly due to rangeproofs.
Your block full of 20K tx would take a monstrous 400MB, and a grin full node
would have need a bandwidth of over 3456 GB / day.
Even a minimal client keeping only the 32 byte transaction kernels
(which can't be pruned) would need storage space of 2TB per year.

regards,
-John
-- 
Mailing list: https://launchpad.net/~mimblewimble
Post to     : mimblewimble@lists.launchpad.net
Unsubscribe : https://launchpad.net/~mimblewimble
More help   : https://help.launchpad.net/ListHelp

Reply via email to