Thanks - some interesting ideas - some comments
Changing from 5 grants to 1 grant likely will not solve anything. That will probably be gamed into an interest pool too.
With the 5 grant system I think coin owners were happy to try to grab one, hoping other owners would support the community grants. But as other coin owners saw the unfairness of this, they wanted in on it too . . . soon anybody who supported the community grants was feeling like a mug, because they were the only ones not getting any interest. I don't think this will happen with a single grant system - either coin owners will support a foundation type grant or they kill the coin. I think there might be disengagement (regrettable), and maybe revolution (fine), but not interest payments.
A long way to solution is first of all massively slowing down the grants - to give a reasonable timeframe for drawing up proposals, laying them out, users understanding them, voting on them.
I was hoping this would happen, but it didn't and I don't see any way to encourage or incentivize it at this point.
Maybe when then coin is more mature - but how to get to that point?
To stop the exchanges or largest holders from dominating the voting, you could cap the vote power at some moderate amount (e.g. max vote is 100 coins), or make it a logarithmic scale. For example 1 coin = 1 vote, 10 coins = 2 votes, 100 coins = 3, etc. Or make a “House/Senate” system where the decision has 2 components: the proportional coin amount as currently, plus a new component of 1 address = 1 vote.
This would counter the self-grant pooling, because the primary motivators for such pools would likely want to divide the proceeds on a proportional basis; small holders would not benefit in such pools, and thus vote for something else, and on a log scale their small vote is more significant. (This might still be game-able - for example, by mass-splitting a wallet, but hopefully that would be more trouble than it's worth?)
Yes, pretty sure anything that adds a cap will cause wallet splitting, and be gamed that way. Maybe not worth it presently, but as the coin grew, this would become the intractable problem.
A 1-address 1-vote component could also have some interesting uses for “American Idol” style votes -- even on issues completely separate from the coin itself. This could generate high interest and demand for MemoryCoin. i.e. General popularity votes, with repeat voting allowed as a “measure of enthusiasm”, yet constrained by the non-zero cost. Consider 2 million Idol viewers seeking 0.10 MEG accounts to vote (!!)
That would be an interesting feature for a coin . . . however I generally regard those tv shows and votes to be pretty obnoxious (stealing money from the young). I don't think there is a way to use it for useful votes because of the wallet splitting problem though. There's no way to enforce one person, one vote. All that works is 'one coin, one-vote' which is proof of stake voting, only really useful for issues related to the coin.
You could set a minimum percent of the total outstanding mint needed for a vote to succeed. Instead of just beating all the others, it has to both win and be above N%. So proposals need widespread community response to win. Grants that don't reach the minimum are not allocated and the funds carry forward to subsequent rounds.
Maybe votes should automatically reset after each round? (This should be a relatively easy code change too - just count the voting transactions over the last N blocks, instead of forever.)
Think these would require a greater level of understanding and engagement with the voting - not sure how to encourage that yet.
Can MEG support multiple styles of voting -- all the above and more, encoded in some way, like in the grant addresses. There should also be votes de-coupled from money awards (this is possible with the current system, but could be made easier to use - so the whole community, or subsets of users could be running their own distributed votes on lots of issues simultaneously).
Think a coin could be used this way as a method of voting - it would be an interesting coin. You'd need an authority to validate users as individuals, unless you were happy to use proof-of-stake in a coin to be a suitable weighting.
I had been thinking from early on, that maybe some grant addresses should be a scarcer resource, or subject to a level of pre-validation. Say they had to be issued and digitally signed by the MemoryCoin Foundation at (small) fee, with some preliminary statement about the purpose. That would give a better record of what the proposals involve, and potentially deter self-grants. Plus it's an additional source of funds for the MCF. And it's reasonable to have the coin foundation enforcing the one rule that “you can't wish for more wishes” (i.e. no self-grant votes)
Trying to avoid any type of centralization. I think crypto-currencies need to be immune from shutdown, so must avoid a required central authority. The main reason for having voting at all is that MCF can be removed by vote when something better comes along.