I prefer deducting a small amount from the each reward. Of course if the resulting payout ends up below zero (or probably still below some dust threshold defined for technical reasons), then the reward would not be paid.
This has the advantage that it can't be trivially circumvented by buying a "top off" vote from a bot (adding even more bloat to the blockchain) and it also better recovers the "costs" of adding content to the blockchain. Finally it is arguably more fair to stakeholders of all sizes, since all must incur the cost. With just a minimum payout threshold, a stakeholder with enough SP to exceed the threshold on his or her own incurs zero cost from it. This is vaguely similar to the situation with n^2 where whales with sufficient SP could reach the very superlinear portion of the reward curve on their own and monopolize influence, while smaller stakeholders could not.
I believe something like this is scheduled for HF20 but I don't recall if the per-content deduction was implement or just the per-vote deduction (the later replaces the minimum dust vote threshold and discourages spam voting in addition to spam content).
One last thought. I believe that an actual 'cost' in the form of a reward deduction (which isn't eliminated once the post reaches a threshold and can't be circumvented) would likely be more effective than a threshold. Therefore to accomplish the same degree of spam control, it could be smaller, all else being equal, and direct impacting small SP holders less. I don't know this for sure, it is just an opinion. Could be wrong.
I agree, that is probably a better way to implement it.
The change that was implemented for HF 20 subtracts a small amount from the individual votes. This is very similar to what you are suggesting, although at the level that they are implementing it - it is probably too low to make any significant difference with spam.
https://github.com/steemit/steem/issues/1764 https://github.com/steemit/steem/pull/1934
It shifts the incentive for people with swarms of small accounts to delegate to either one voting account or a paid delegation service rather than have each account vote with a bot. That may not affect content spam but it does reduce vote spam (in the case of delegating to a paid service, it may even reduce content spam, with many variables). Although for people with small accounts receiving delegation, they won't be able to do that (since redelegation isn't possible), so the problem persists. At least in that case a possible solution is getting the delegation removed.
Nuff said here and now, Thanks for your time.Hello @smooth I recently became aware of huge issues with fake accounts going back to day 1. I see now that many good people are aware of this and trying to implement corrective measures. @anonymint pointed me in your direction. Is there some off chain area where info is posted? Like who are the good guys VS the bad? I would like to at least use my witness votes correctly. Bad actors from day 1 still have immense control and power here.
I don't know of a specific resource to suggest. Be engaged both on the site and on steem.chat, learn, and reach your own conclusions about "good guys VS bad". This is not an easy problem with a cookie cutter solution.
This could be taken to @anyx, @patrice, and company at http://steemcleaners.org for reporting, but they have to chose their battles by severity and applicability to their mission to fight spam and bot rings. But it's a resource to at least consider. There is also #abuse I think it's called, on http://steem.chat but that's a pretty noisy channel.
Finally! Someone who mentioned HF 20. I'm including a screenshot of the two paragraphs from the late December update by steemitblog that address the changes that are set to occur when HF 20 drops. It not only talks about a 1.219 shift down of SP, but the elimination of the dust threshold.
It also says that basically powerless votes will have no impact on the blockchain, just I presume, as it is now. Just that the user will be able to vote to their hearts content.
So, if this is still happening, and the dust threshold is eliminated, how can it be raised (this question is for you or @timcliff or anyone else that can answer it) if it doesn't exist? And wouldn't this and the idea of raising a non-existent dust threshold be working at cross purposes, if a no defunct dust threshold could be raised?
Please excuse my confusion. Maybe the problem is the terminology being used, ie, dust threshold, has two different meanings here: one affecting the ability to vote, the other affecting if someone can be paid. As it is, I think both are addressed in the screenshot above, so I'm still not sure how this raised dust threshold is supposed to work. Add in the SP shiftdown, and I'm even less sure.
AFAIK that vote shift stuff is still in the HF20 development code.
On the topic of dust votes and dust rewards (as well as this post), there are two different issues, one concerns individual (very small) votes and the other individual (very small) reward payouts. I'm not aware of any already-implemented changes to the dust rules for reward payouts in HF20, only for dust votes.
Okay, that's the distinction I was missing dust votes vs dust payout. Thank you for clearing up my confusion. I really do appreciate it.