Why oh WHY do people do stuff like this!? I know that this was already
discussed somewhere else, but the block rewards equation is rather difficult to plot/understand, and it really should be clarified.
Block reward: Smoothly varying using the formula (264 - 1 - A) * 2-20 * 10-12, where A = supply mined to date
Assuming that's correct, if we want to break that down into something people are actually able to understand (without resorting to serious spreadsheet calculations or function plotting software), let's start by removing those exponentials:
Block reward...
= (2
64 - 1 - A) * 0.00000000000000000095367431640625
= 17.592186 - 0.00000000000000000095367431640625 - A * 0.00000000000000000095367431640625
= 17.592186 - 0.00000000000000000095367431640625 - A * 0.00000000000000000095367431640625
If A is 0 to 18400000, we could basically drop both the second terms and the block rewards would be ~17.592186. Except that's clearly not the case, so what's going on? Well, A is obviously not going from 0 to 18400000; it's instead going from 0 to some large number -- but not so large that (A * 0.00000000000000000095367431640625) would be more than 17.592186, since we don't want negative block rewards.
I looked around in the source code (which is frankly a mess to understand -- hooray for security through obfuscation and complex code!), and anything even remotely resembling "2
64 - 1 - A) * 2
-20 * 10
-12" is nowhere to be seen in the block rewards calculations. Instead, we get bitwise shift operations and other items that again only seem to serve to obscure the true meaning of what's happening. It appears the coin supply (A) is actually scaled by 1,000,000,000,000 (one trillion) -- the reward is also limited in some way by the block size (
). So, applying that scaling factor, what we get is that the block rewards are
17.592186044415 - A * 0.00000095367431640625, with A going from 0 to 18400000. (I dropped the second term as it appears to be insignificant -- even when scaled by 1 trillion.)
The result is that we're seeing block rewards of around 17.14 right now, so we're down ~0.45 from the initial rewards level in a couple weeks I guess. Over the long haul, it looks like we smoothly scale block rewards such that the reward is cut in half every ~504 days (give or take) -- and about
half of the remaining coins are mined every ~504 days. That's really not that bad, but rather than such a clearly stated progression we get "Block reward: Smoothly varying using the formula (2
64 - 1 - A) * 2
-20 * 10
-12, where A = supply mined to date".
Thanks for that... it only took me a few hours to actually dig through the code, run some numbers, and figure out in plain English what is actually happening. It will take about nine years for 99% of MRO to be mined, or 4.5 years for 90% of MRO to be mined. That's a far faster distribution than BTC or LTC of course, but far slower than most of the other coins coming out these days.
Back to "real" work.... LOL
