This is perfect, if it works for MrData... I can just reference this chart in the OP and webpage.
Is this a new concept, or have other coins done it? You should name it if it's a new concept. Do you want to call it "smooth reward reduction"?
Topic: [ANN][SPT] - Spots - Cryptsy | Fork 9/19 | Scrypt | Introducing Reward Reduction - page 13. (Read 93329 times)
full member
Activity: 224
Merit: 100
"Change is the only constant." -Heraclitus
September 10, 2014, 02:54:30 PM
September 10, 2014, 02:04:44 PM
Played a little with numbers and decided that good combination will be 0.95 (95%) of previous reward each 48k blocks with ceiling to 0.2. First graph is reward after block #, second graph is reward after passing of #days (beginning from reward 24spt/blk).
As you can see from this pics is that 12.2spt/blk (~half of 24spt/blk) will be at block #946'000, after 429 days from first 24spt/blk.
6spt/blk will be at block #1'858'000 - 971 days from first 24spt/blk.
And minimum block reward will start at block #2'386'000 after 1286 days from first 24spt/blk and will never decrease.
If you want .xls of this data - write me pm with your email, i'll send it.
As you can see from this pics is that 12.2spt/blk (~half of 24spt/blk) will be at block #946'000, after 429 days from first 24spt/blk.
6spt/blk will be at block #1'858'000 - 971 days from first 24spt/blk.
And minimum block reward will start at block #2'386'000 after 1286 days from first 24spt/blk and will never decrease.
If you want .xls of this data - write me pm with your email, i'll send it.
September 10, 2014, 12:43:42 PM
What minimum reward do you want? (ie. so the reductions never take the reward below X amount.)
If halving every NNNNNN coin (f.e. 11 mill) mining take so time:
48 now - 1 year 2 month
48 > 24 ~1 year 1 months
24 > 12 ~2 years 2 months
12 > 6 ~4 years 4 month
6 > 3 ~8 years 9months (!!!! Really?)
Lifetime more 14 years!
Below 3 coins per block seems to me uninteresting
September 10, 2014, 12:36:14 PM
What minimum reward do you want? (ie. so the reductions never take the reward below X amount.)
Edit: Also how long do you want before the hardfork should kick in?
Edit: Also how long do you want before the hardfork should kick in?
full member
Activity: 224
Merit: 100
"Change is the only constant." -Heraclitus
September 10, 2014, 12:02:44 PM
This algorythm will prevent sharp reward drops. And this of course will prevent situation when most of miners will out when "next" halving occurs. Decreasing of reward in this algorhytm will be semi-smooth and will better effect at mining and trading. And i don't think it's so hard to program such algo. Once i saw bitcoin source and read this reward halving block. That was a simple code. Adding some formulas will increase code by 10 lines maximum.
I'm not suggesting it is hard to program. MrData commented earlier that it was something he felt comfortable accomplishing. I am suggesting that to explain it in the OP, and for SPT users to understand, it is complex to understand.
So, in general, what you are suggesting is that the block reward drop by 5 - 10% instead of 50%... In this scenario you would do the reduction more frequently than your standard halving. The goal being to not significantly change the mining / trading climate by drastically reducing the coins being placed in circulation.
MrData - I don't think this makes a big difference either way. I think Matory and I are comfortable with whatever is decided upon. No one else has chimed in. Please move forward how you feel most comfortable, since the majority of the work will fall on you.
September 10, 2014, 11:13:30 AM
This algorythm will prevent sharp reward drops. And this of course will prevent situation when most of miners will out when "next" halving occurs. Decreasing of reward in this algorhytm will be semi-smooth and will better effect at mining and trading. And i don't think it's so hard to program such algo. Once i saw bitcoin source and read this reward halving block. That was a simple code. Adding some formulas will increase code by 10 lines maximum.
full member
Activity: 224
Merit: 100
"Change is the only constant." -Heraclitus
September 10, 2014, 09:20:30 AM
Desten - If you don't reply by later this afternoon, we'll have to move forward in another direction. I appreciate your opinion, and don't want you to think it is being disregarded. What you are proposing is complex, and I don't understand the benefit of doing it this way.
full member
Activity: 224
Merit: 100
"Change is the only constant." -Heraclitus
September 08, 2014, 08:24:28 PM
Desten, I think what you are proposing would be difficult for your average crypto-user to understand. I think it would cause unnecessary confusion.
What is the benefit of doing it this way? If the benefit is significant, then maybe we can work beyond the risk of confusing users.
What is the benefit of doing it this way? If the benefit is significant, then maybe we can work beyond the risk of confusing users.
September 08, 2014, 02:12:07 PM
The simplest is halving by block count, but the 95% one isn't very hard either.
September 08, 2014, 01:39:59 PM
I'm confused in the case of simple make harder. Notify me when a solution is found.
September 08, 2014, 12:56:49 PM
Halving after 11 mill and decrease to 0.95 (0.90 or 0.92) every 11 mill or every 1 year? 11 mill is better.
If it is doable and is not difficulty - let's decide this!
If it is doable and is not difficulty - let's decide this!
No. Halving @11-11.5 mil because it's simplest thing now and will give some time to update wallets at exchanges.
After that i think reward algo must be changed to smooth decreasing.
For ex.:
0-11mil spots: 49/48 (before and after first fork). This will be ~226k block.
Here are two examples, first is 95% after each 16k blocks with ceiling to 0.25 (after 224k blocks next reward will be 24.75 spots, little more than a half of 48 spots block, next "halving" will be after 256k blocks), second is 90% after each 32k blocks with ceiling to 0.1 (after 224k blocks next reward will be 23.2 spots, next "halving" will be after 224k blocks) and so on. This is just example what can be if this algorithm is used. But we decided to divide block reward soon, so this calculations will start from 24 spots/block. Because of ceiling (to 0.25 up or to 0.1 up in this examples) after some time reward will not be decreased and will flow constantly
| Now | Reward after decr., ceil(95% of prev., 0.25) | Decreasing after +blocks | Now | Reward after decr., ceil(90% of prev., 0.1) | Decreasing after +blocks | |
| 48 | 45.75 | 16 | 48 | 43.2 | 32 | |
| 43.5 | 32 | 38.9 | 64 | |||
| 41.5 | 48 | 35.1 | 96 | |||
| 39.5 | 64 | 31.6 | 128 | |||
| 37.75 | 80 | 28.5 | 160 | |||
| 36 | 96 | 25.7 | 192 | |||
| 34.25 | 112 | 23.2 | 224 | |||
| 32.75 | 128 | 20.9 | 256 | |||
| 31.25 | 144 | 18.9 | 288 | |||
| 29.75 | 160 | 17.1 | 320 | |||
| 28.5 | 176 | 15.4 | 352 | |||
| 27.25 | 192 | 13.9 | 384 | |||
| 26 | 208 | 12.6 | 416 | |||
| 24.75 | 224 | 11.4 | 448 | |||
| 23.75 | 240 | 10.3 | 480 | |||
| 22.75 | 256 | 9.3 | 512 | |||
| 21.75 | 272 | 8.4 | 544 | |||
| 20.75 | 288 | 7.6 | 576 | |||
| 19.75 | 304 | 6.9 | 608 | |||
| 19 | 320 | 6.3 | 640 | |||
| 18.25 | 336 | 5.7 | 672 | |||
| 17.5 | 352 | 5.2 | 704 | |||
| 16.75 | 368 | 4.7 | 736 | |||
| 16 | 384 | 4.3 | 768 | |||
| 15.25 | 400 | 3.9 | 800 | |||
| 14.5 | 416 | 3.6 | 832 | |||
| 14 | 432 | 3.3 | 864 | |||
| 13.5 | 448 | 3 | 896 | |||
| 13 | 464 | 2.7 | 928 | |||
| 12.5 | 480 | 2.5 | 960 | |||
| 12 | 496 | 2.3 | 992 |
September 08, 2014, 12:37:42 AM
And why you all rejecting smooth division by 5-10% of previous reward?
I've got no problem with it.
Halving after 11 mill and decrease to 0.95 (0.90 or 0.92) every 11 mill or every 1 year? 11 mill is better.
If it is doable and is not difficulty - let's decide this!
September 07, 2014, 08:57:32 PM
And why you all rejecting smooth division by 5-10% of previous reward?
I've got no problem with it.
September 07, 2014, 05:56:27 PM
This will lead to another problem in future if dividing will be each 11mil spots. Divide by blockcount.
And why you all rejecting smooth division by 5-10% of previous reward?
And why you all rejecting smooth division by 5-10% of previous reward?
September 07, 2014, 05:06:33 PM
I just mean the 1st halving would have been 6 months if the blockchain had been running smoothly, then doubling after that.
September 07, 2014, 04:56:37 PM
Just FYI halving every 11 million coins would be about every 6 months; the only reason it's taken a year to get this many blocks is because of the old v1 diff algorithm that seized the blockchain up so much.
First 11 million - ~1 year 2 month
Second 11 million - 11,000,000/24=458333 blocks
Or 458333blocks*70sec/60sec/60min/24hour ~371 day
Third 11 million ~ 742 days or 2 years 2 weeks
......
September 07, 2014, 04:51:02 PM
Just FYI halving every 11 million coins would be about every 6 months; the only reason it's taken a year to get this many blocks is because of the old v1 diff algorithm that seized the blockchain up so much.
September 07, 2014, 04:48:41 PM
I like how Spots working now. Halving block is the best solution for me. I propose insert half block every 11 million coins.
With 1block=24SPT next half is after ~1 year from 11 million coin date.
With 1block=24SPT next half is after ~1 year from 11 million coin date.
That's certainly doable and if we go by our current block reward rate not accounting for the early v1 reward structure would give us about 11-12 days until the hardfork. If we do account for the old reward structure it would be about 9 days (~3000 blocks difference.) and add slightly to the code complexity.
I'm up for it. The timing looks to be pretty good as well, giving us about 2 weeks to get clients updated. I like the idea of a slow halving, around once / year. I think the fact that we are doing it, will be good for the market.
Does anyone else have thoughts on Scrypt-N vs another algo? It sounds like the consensus so far is to leave it alone...
Vote to keep Scrypt-N (or X11 if possible)!
full member
Activity: 224
Merit: 100
"Change is the only constant." -Heraclitus
September 07, 2014, 04:36:34 PM
I like how Spots working now. Halving block is the best solution for me. I propose insert half block every 11 million coins.
With 1block=24SPT next half is after ~1 year from 11 million coin date.
With 1block=24SPT next half is after ~1 year from 11 million coin date.
That's certainly doable and if we go by our current block reward rate not accounting for the early v1 reward structure would give us about 11-12 days until the hardfork. If we do account for the old reward structure it would be about 9 days (~3000 blocks difference.) and add slightly to the code complexity.
I'm up for it. The timing looks to be pretty good as well, giving us about 2 weeks to get clients updated. I like the idea of a slow halving, around once / year. I think the fact that we are doing it, will be good for the market.
Does anyone else have thoughts on Scrypt-N vs another algo? It sounds like the consensus so far is to leave it alone...
September 07, 2014, 04:26:14 PM
I like how Spots working now. Halving block is the best solution for me. I propose insert half block every 11 million coins.
With 1block=24SPT next half is after ~1 year from 11 million coin date.
With 1block=24SPT next half is after ~1 year from 11 million coin date.
That's certainly doable and if we go by our current block reward rate not accounting for the early v1 reward structure would give us about 11-12 days until the hardfork. If we do account for the old reward structure it would be about 9 days (~3000 blocks difference.) and add slightly to the code complexity.
Jump to: