Hardfork 20 - missing information
Firstly, I have upgraded to 0.20.1 (Edit: on 0.20.2 now) and support Hardfork 20. I won't be satisfied till there are radical changes (like replacing stake-weighted voting), but this is a step forward in every way. With HF20, witnesses have to make new decisions regarding subsidized accounts - budget for subsidized accounts per block, and decay of subsidy pool per block. However, I'm unable to make any decision without more information. I (and others) have requested this information twice on Steemitblog posts, only to be duly ignored. I'll ask again, and will greatly appreciate any responses from people privy to such information. By the way, why should a middling ranked witness care? In cases of divided opinion between the top 20 witnesses, parameters change every block as per the runner-up witnesses.
Historical charts for
a) New account sign-up requests from Steemit.com and elsewhere.
b) Percentage of rejected sign-ups.
c) Percentage of accepted sign-ups which turn out to be spam accounts.
d) Percentage of total registrations processed by @steem (I do note that in all these cases, the subsidized account system has the potential to radically change how registrations are done, deeming the historical data irrelevant. Nevertheless, it's a good place to start.)Steemit Inc's intentions -
a) Will Steemit Inc participate in the discounted account market? If yes, to what extent? They have an utter monopoly on RCs to bid with, given their exorbitant stakeholding. This is an important question that needs answering.
b) "To what extent" has many possibilities - They could use the subsidy pool, but for each account created, burn Steem for most or a part of it. Or, commit to use only a part of the subsidy pool, while fully burning Steem for the rest. Or combinations thereof. I need to know precise details on how Steemit.com will handle account sign ups going forward. Of course, all of this will be evident after HF20, but some heads-up feels necessary.
c) Are we aware of other entities that intend to participate in this market?Commitment to spam combating measures by Steemit Inc's faucet, and other third parties. (Remember when social networks could just ban spam accounts?) Obviously, I'm not asking for their method, but I want to be assured there will be significant spam mitigation protocols if witnesses do choose to open up registrations en masse.
Simulation of historical blockchain usage and how it will translate to the new RC system. I'm expecting teething issues now that each transaction has a different cost, but I have no idea how it'll play out.
Historical spikes and trends in account sign-ups, which will be vital information for decay rate. Of course, this will be broadly correlated with Steem price (pretty much every activity metric on Steem is), so I anticipate adjustments with significant price swings.
Any further information would help, but these would be a bare minimum. If I don't receive further information, I'd err towards the side of caution, and go with a low budget (perhaps <1 per block), an average decay rate (5 days), and a high fee (~12 Steem at current prices). If 1d) suggests that Steemit Inc is creating an overwhelming majority of accounts, them burning more Steem is ultimately best for the ecosystem, and ties in perfectly with Steemit Inc's stated goal of reducing their stakeholding over time. Or so was stated in the 2017 Roadmap, not sure if that has changed. On the other hand, this data will not account for those planning to jump into account creation as a result of the subsidized accounts. Anyway, I'll of course revise these numbers the first week of October once we get more information about how Steemit Inc is using these new protocols, in practice.
Some of this information can be scraped from the blockchain, and I'm sure it would be of great interest to all witnesses and stakeholders. Ideally, the top witnesses should hire researchers and get independent reports made for a more reasoned perspective. Other stuff, well, we'll just have to hope we get something from Steemit Inc.
Thank you to all developers for getting us to Hardfork 20, and for users of the platform for being patient. See you on the other side of HF20 next week!
PS: I'm trialing power up 100% because SBDs may start printing again post HF20. When SBDs are under $1, it's beneficial to go the power up route, if you were intending to power up anyway. In the past, this also came with social capital and appreciation by the community, though that has since eroded. However, do note that you stand to lose out if in one week's time SBDs shoot well above $1. On a related note- in the face of unstable crypto traders and rampant manipulation, I'm not going to publish an interest rate or a bias if SBD trades in the $0.90 to $1 range.
PPS: This is a good analysis to start with - https://steemit.com/steem/@paulag/exploratory-analysis-account-creation-and-hf20-changes-1537128103339
@paulag has recently collected some data on the account creation stats - her post and interative visualization may give at least some information on the points you addressed. Most of them require insights from Steemit themselves, I guess....
thanks for the mention and shout out @corkkon
@liberosist it seems most of these discussion are not being held on the block. Lets hope they are being held somewhere because there is a lot to think about here. As a lower level witness (@steemcommunity) I would be very interested in hearing opinions on this. Personally right now, I would keep the parameters as low as possible until there is some communication from the top witnesses.
I would love @ned and steemit inc to share some data with us. But heres hoping :-)
Thank you, your recommendation is much appreciated. I'll go through it in detail soon.
As far as I know Steemit intends to use the subsidized account creation to the maximum extent possible. I think we need to be very careful about the rate of subsidized account creation. Each new account 'comes with' a specified amount of 'free' RCs which means, going forward, increased ongoing demands on the system without any increased demand for STEEM (or, viewed alternately, a dilution of available RCs for existing STEEM/SP holders).
Remember that they are already creating "free" bandwidth without increased demand for steem, as the steem is ninja mined.
Arguably true but there is an inherent multiplication with these 'free' accounts. If you use X RCs to create an account, that new account then throws off Y RCs per day forever. If the ratio between X and Y is too low this can have a profound effect on the system.
Thanks. As long as Steemit Inc is the dominant beneficiary of the subsidy pool, I completely agree we need to be very careful (hence this post). I'll be publishing a very low budget to begin with, and see how things progress from there.
Without this information from STINC, I will likely set the lowest parameters possible until reliable new data is gathered. Their carelessness and lack of cooperation on these matters tells me that we’ll continue seeing the same amount of incompetence and the same repeated mistakes that waste network resources and contribute to the lackluster performance and perception of their site and company...which reflects poorly on the rest of us here.
Lowest possible is... 1 subsidized account per 10,000 blocks. Which is - ~3 subsidized accounts per day?
I would have no qualms about this either, if no one else from the community (developers mostly) was intending to take advantage of the subsidy pool. If it was going to be Steemit Inc, Blocktrades and Anonsteem dominating account registrations as usual, it'd be best to make them burn for it. However, if a developer comes up with an impressive plan to market their apps, I'd like to see them get access to at least some subsidized accounts.
But yes, that seems far fetched at this stage, because we haven't seen a single developer on Steem with any marketing plan whatsoever in the past 2.5 years.
I'd support removing stake-weighted voting, but how would it be done? If every vote was equal, what would stop someone like noganoo from ruling over the community with his millions of scam accounts?
That's what the hold up is for me.
Unfortunately, I don't have a solution. Ned has been hyping up account-based voting (and trashing stake-weighted voting while at it; which is weird because that's still how the Steem reward pool works with no plans announced to change it), where accounts are whitelisted. To be fair, account-based voting is something that might work for smaller scale, centralized SMTs. Alternatively, SMT moderators can blacklist, so the noganoos of the world would be blacklisted from allocating rewards, which could be more scalable if done right. Smart reputation systems could work too, if someone cracks that.
Of course, account-based voting of any sort is not going to work for the common Steem reward pool. My only thought is just remove the common Steem reward pool, and let the SMTs and Communities compete for the best solution. Users will flock to the SMTs that figure it out. You may remember I ranted about this a few months ago, and haven't found a better solution since.
The common reward pool has devolved into a shitshow of scamming and abuse, and no one has a solution in sight. It's discovering new bottoms of the barrel every week, and nothing's been done about it. I don't have anything else to add...
That’s because Ned has no earthly idea what he’s doing. He has called Steem protocols “evil” and trashes the very system that he and his “dev team” reside over. The better option for him - in my opinion - would be to power down and leave. Instead, he’s intentionally breaking a system that had coherence, but now has none.
Whether or not you liked the original system, it at least made sense. What we have now is a bunch of mismatched pieces of a puzzle with no picture.
There are some solutions to the reward pool waste/abuse that we see. But nobody is willing to try them because the handful of large stakeholders and vote-schemers are making too much easy money now...which is why the protocol changes last year should not have been made. They were far more “evil” than what preceded them.
I don't agree that the system was ever coherent. I mean... hyperinflation and n^2 were batshit insane. They are so nonsensical that I often wonder whether Dan only ever came up with those to increase his own stake so he could dump and leave. But yeah, it's become a complete mess now.
I was never a fan of the hyperinflation. The coherence I’m talking about was with the reward allocation protocols for content and curation. I think we all agree that n2 was too steep. But today’s Frankensteem is a much worse horror show.
Full linear is a problem. The 10-vote target is a problem. The “75/25” reward split is a problem. (I would also add SP delegation to this, but others disagree for some reason.) Together, they create an incentive structure that is not aligned with investor interest and that provides poor results for social media user interest. And more proposals are on the way that will exacerbate some of these misalignments.
The entire ninja-mine + hyperinflation was nothing but a power/money grab. I think that’s the only true consensus on this platform.
BTW...I haven’t seen Ned try to fix that issue yet. Not even once. In fact, STINC continues to use that stake to seemingly line their own pockets and the pockets of their friends (and scammers and BS “projects). So the next time Ned talks about “evil,” ask him about that. (Don’t expect a meaningful response.)
Agree with you here
^2 is too much but some sort of superlinear is necessary. What linear boils down to is rewarding content agnostic behavior the most. Whether it's self voting or vote selling, the most direct way to maximize profits basically disregard the content entirely. And no, even if bots impose standards, they'll be out-competed by those who don't. Even if not, minimum standards are not good enough, it's basically a quality cap, with no incentive to go beyond that.
We're left with a mess of a platform that rewards and displays content that has no reflection or bearing of quality whatsoever. Imagine if facebook, twitter, reddit, hell even itunes, spotify etc. just threw any content onto the front page/feed. They'll be no incentive to visit them at all. Yet if stakeholders don't self vote or sell votes, they just lose stake relative to those who do under the current economic system.
SMTs are great and all, but they alone cannot solve this if the underlying token of the chain (Steem) is so dysfunctional. A few popular SMTs can't carry the main token. Removing stake based voting also takes away the biggest incentive to hold a stake as well.
Have maybe n^1.3 with 50/50 curation and maybe 10% of your total voting power in free downvotes a day. I honestly think this would be sufficient to turn this entire place around. Voting on good content (at least subjectively good content) and competing over curation will likely out compete self voting/vote selling under this system. This platform will be a functional content discovery platform again, and they'd be more incentive to buy steem if the voting market isn't so rampant.
I'm so surprised our voices here have fallen on to deaf's ears for so long. I don't think it helps attacking Steemit Inc or Ned, we really need their help here, along with the witnesses. I'll try bringing this up with them again (i've failed a few times before) after the dust has settled from this hard fork.
Most likely.
For some added context on what I’ve mentioned here...
https://steemit.com/steemit/@ats-david/make-steemit-great-again-fork-this-place
Remember when we all talked about this stuff and wanted to find real, scalable solutions? What happened to that? When did all of that change?
(Hint: shortly after last year’s hard forks)
Your #2 is already implemented and #4 and #5 are included in HF20. That's not everything you asked for but it seems like significant progress to me.
Ah yes, I remember that post!
I like stake weighted voting. Have you tried Trybe? I think they have the kind of voting you like but i like the Steem voting better.
I think, the blockchain was down for some time because of some issue in the HF20, is that now addressed and we are moving ahead with HF 20 ?
Yes, that has been fixed and HF20 is going ahead on September 25th.