Hey peoples. Just in case you haven't caught on yet, I will be helping VanityWallets2015 with the technical aspects of the fork to ensure everything goes smooth. A couple suggestions:
I recommend raising the targeted block spacing from the current 40 seconds, to something like 120 seconds. This is a very small price to pay for significantly less inflation and in my opinion is definitely worth considering. It will also add a larger queue of blocks that are ready to stake, which will increase the PoS difficulty (safer network) and also ensure that the chain always has coins ready to move transactions along.
I also recommend reigning in the time drift to 2-3 minutes. This is a technical matter and if you would like to view the reasoning see
this interview I had with bitcoinist.net
And I also wanted to let you know that I am willing to accept ANAL as well as BTC. ANAL I would price the value at the 7,000 sats. I started an address here AeaMQ4EuaEpgG1UpqUEeLsiGgawuANuuBW for the fork donations. I am excited to be working with you all, and have purchase a few coins off of the exchange to get ready to stake
I would support measures to make the chain more robust. Block time of 120s however is long for retail especially if you add the confirmations. Inflation is something that can be sorted out with a future monetary policy review.
In order to compare the options, one has to calculate the interest compounding by stake. For a 40s block time there will be 788,400 blocks in one year on average and for 120s block time there will be 262,800 blocks in a year on average. This works out to 492,818 coins after one year 40s stakely compounded and 492,815 coins 120s stakely compounded (3 coins difference). Compare this to the graphs above where the total worked out to 490,089 coins daily compounded after one year (±2,000 coins difference). Unless my calculations are way off and made a mistake and someone can post a better estimate, the block time seems to have immaterial impact on inflation. Unless a longer block time has a material improvement on the robustness of the chain, I would recommend keeping it at 40s or perhaps even make it 30s.
I am pretty sure your calculations are off, because you cannot really measure the difference in compounding. Because a longer block time means a longer period it will take to compound because you are less likely to stake right away.
A small block time, although sounds great, is a great way to bloat your block chain really quick, which leaves users one or two years from now having a very difficult time syncing.
I am not the one to make the decision for you all though. Just a simple recommendation from what I have seen out there in the PoS world
Excellent interview and explanation of time drift. Agree with your suggested specs for time drift.
Regarding block time. The compounding period is subject to the law of diminishing returns and the total % of POS interest is fixed to 125% nominal interest if I understand POS design correctly. At some point the number of compounding periods have a very small impact on the total. Initially the impact is greater but later on it does not matter if you use 100,000 or 1,000,000 compounding periods. See graphs below that is based on the total population of coins that therefore own all stakes and a starting number of coins of 141,195. However, I understand the impact of number of blocks on the size block chain. BTC is 20+GB over its life, although it must be said it depends on the number of tx in a block. Having said this, in the case of BTC there might be 1,800 tx in one block during 10min, but with a 60s block chain it might have had 180 tx per block so the net increase would be the fixed bits per block excluding tx. If the fixed part is large then yes there would be a much more bloating. I think the answer is in the intended use. Will the coin be used more for just paying on a website or will it be used for day to day retail payment? If the intended use is not block time sensitive, then a longer blocktime can be used. If the use is time sensitive, then a shorter block time must be used.
I would still keep with the 40s and re-evaluate in a year or so if the decision is based on inflation of coin alone. It is a different matter if the decision is made on bloating of the blockchain. Any other views from the community?
Up to 100,000 compounding periods in a year and for that matter up to 788,400+
Up to 365 compounding periods in a year
Up to 12 compounding periods in a year
EDIT:
I just had a look and compared the experimental 10K blockchain size and BTC blockchain sizes at approximately the same height of 360,000 blocks. 10K is around 400MB and BTC 20+GB. This seems to indicate that the tx content of the blocks contribute much more to bloating than the fixed overhead since 10K is pretty empty tx-wise.
)
