Why?
Without the change roughly 99.9999999% of the 21M BTC will be mined. Higher precision will require larger bit length to hold values. This means larger transactions, more cost in storage, bandwidth, etc. There is no reason to increase precision until 1 satoshi becomes a constraint on commerce (say $0.001 USD circa 2012).
So to summarize.
The cost:
A hard fork
Which requires constant updating
Increases the size of all future transactions
vs
The "benefit":
Money supply is increased by at best 0.2% (the amount of that 0.2% which can be "captured" will depend on the number of digits added).
However PSA:
All future crytpto-currency inventors please ensure if you use a subsidy and the subsidy declines by right shifting make the initial subsidy a power of 2. This will result in the subsidy cleanly going to exactly zero at the point all coins have been minted. Also make the subsidy interval a multiple of the difficulty interval.
Topic: Delay BlockReward=0 Forever - page 2. (Read 1953 times)
Dear Technical Users,
Can we simply require that more decimal points be added if the block reward were to drop to 0 (or a few years in advance of), so instead reward eventually becomes .5 satoshi (0.000 000 005 BTC, at this point divisible to 9 digits), then .25 satoshi (BTC divisible to 10 digits), .125 (BTC divisible to 11 digits), etc?
http://xkcd.com/994/ (Happy Holidays everyone)
I know 'adding decimal places' is a hard fork, but if we forked it this way (to constantly expand on schedule) we might *only* need to fork it once.
...
Now tell me all about how you hate this idea.
Can we simply require that more decimal points be added if the block reward were to drop to 0 (or a few years in advance of), so instead reward eventually becomes .5 satoshi (0.000 000 005 BTC, at this point divisible to 9 digits), then .25 satoshi (BTC divisible to 10 digits), .125 (BTC divisible to 11 digits), etc?
http://xkcd.com/994/ (Happy Holidays everyone)
I know 'adding decimal places' is a hard fork, but if we forked it this way (to constantly expand on schedule) we might *only* need to fork it once.
...
Now tell me all about how you hate this idea.
Jump to: