Rejecting HF21 in its current state

in #steem3 years ago (edited)

Firstly, let me be clear that I'm a low ranked witness, and my opinion is largely irrelevant. Nevertheless, I'll do my duty as a witness and state my sincere opinion. If I reject a hardfork that everyone else accepts, it'll mean that I'll be kicked out of Steem and no longer be an active witness.

I support the general idea of the EIP, overall, but I do not think the current implementation proposals being chucked around are justified.

I have discussed the pros and cons of each proposal in the EIP, in general, here.

Addendum: Before commenting, please see my full thoughts in the comment linked above. This post is but a follow-up from there.

However, I'm yet to see any data, simulation, or evidence, to justify the dramatic changes, all rolled into one hardfork. The change to 50:50 is a 100% increase in curation rewards, a 33% pay cut to the content creators without whom there'll be no curation; giving abusive whales 10%-25% free downvotes is drastic etc. Remember, one abusive whale can effective now vindictively kill off hundreds good content creators at zero cost. There's a greater cost for minnows to unite and start a flag war with whales, whilst for whale it's pretty easy to set up a bot and just take a bunch of people they don't like down to $0 forever. Self-voters can simply set up a bot to revenge flag anyone who dares downvote - it'll always be minnows that'll be forced to retreat from a flag war. I'm also concerned about the convergent linear curve hampering the few genuine minnow content creators, but overall I'm in favour of it. Finally, SPS is going to be introduced without any funding committed by anyone. You can read my detailed thoughts in the comment linked above, and alternative brainstorming in my blog.

I was expecting better implementations of the EIP concepts, but these are simplistic and naiive. Have we learned nothing from the past?

To conclude, I remain in support of the general concepts of EIP, but disappointed by the facile implementations proposed thus far. I will, of course, change my opinion based on further tweaks to HF21, and I hope there'll be some.

PS: As some have correctly pointed out, there has been no official announcement about all of this being included in HF21. Nevertheless, my opinion on all of the above is contained here or linked elsewhere. I will, of course, reevaluate my opinion based on the final HF21 proposed by the developers.

PPS: In a follow-up post, I explain why moving to 50% curation rewards will not see any significant changes.


I wish we could run some simulation though. But it’s just not like running a purely engineering simulation to see if a tower or bridge will collapse. With the economy, there needs to be stakes at hand and enough people taking it seriously, which a live environment can only likely do. It doesn't mean we can't attempt to make some intelligent guesses as to what should be done though.

We can apply the measures one by one, but will it achieve the intended effect that was hypothesized in the first place? I don’t necessarily see it as multiple changes by the way, as they’re in effect, just one change. For example, we can simulate the 25% downvote measure and see how things turn out (still not sure how to simulate properly). If we roll out just that.. there are other ways to circumvent the goal of the EIP, hence the other measures.

Personally I'd prefer DVs on the lower side if its delegatable. I think 25% is a yardstick proposed for something sufficient enough to turn the status quo around, accepting that it'll definitely create a lot of toxicity, while being low enough that it can't be used to push rewards in any meaningful way. If we go too low on this then it might be insufficient to turn to this place around to a new equilibrium while introducing similar levels of abuse / toxicity as 25% DV would have, rendering it a bad move.

I've tried to explore the topic more in-depth here:

Good post, it's the most interesting one arguing for EIP that I've read so far. As before, we'll have to agree to disagree, no need to beat the dead horse as we've discussed the matter enough. :) You can see my comment linked above for why I think bidbotting and self-voting will not see a major change. That said, I'll just say this - Games have moderators, designers, judiciaries and referees. In addition to changing the rules in a very timely manner, pay-to-win games work because abusers are banned en masse. Similarly, social networks have teams of thousands dedicated to finding abuse, banning miscreants and implementing countermeasures. Till this fundamental problem is solved, Steem will always be abused and exploited. Thing is, these problems have been solved too, as stated above. So are we going to ignore historical solutions that are backed by overwhelming evidence? In favour of mere guesswork? I get it - these solutions will take time to implement; in the meantime we can try some quick fixes and see how it goes, which is why I'm still on the fence. But we have to start thinking about the big picture.

Some things can easily be simulated (especially now that I've seen the Snax team run simulations); while others are more complex and require deeper psychological, econometric and anthropologic research. Of course, it'll be a combination of both in the end. I'm not expecting complex neural networks just yet, by the way, but as a civilisation, us humans have a rich recorded history to derive knowledge and statistics from. I mean, even from Steem itself, you can study a lot e.g. the downvoting relationships. As you may know, in the early years of Steem, I have written a lot in the past paralleling Steem to political models, historical societies, play-to-win games, evolutionary models, even universal systems! And of course, other social networks. I was mistaken because I thought Steem has a chance of being a new paradigm, so some of the previous learnings may not apply. Yet, through every single parallel I drew, Steem has very predictably been boiled down to everything I said it has a chance of not being. Steem is a society / network of humans like any other, and all of the same predictable rules apply. Anyway, there's no research, no audits, no evidence of any sort published regarding the matter. Currently, the proposals are just guesswork, and it's simply not good enough when a swath of major changes are proposed. I hope there'll be a testnet, at least, before the implementation is frozen and the hardfork is proposed. Sorry for the haphazard rant, I'll end it here!

I can understand that you've got concerns about the effectiveness of the EiP, but doing nothing is not an option and seeing the results of EiP is not possible in theory. At least not completely.

Of course, you can chip in your opinion as a backup-witness, but I'm wondering why your opinion should be valued while you've sold pretty much 99% of what you've earned here on Steem. Isn't this clearly a sign that the current system isn't incentivizing people enough keep their rewards aka Steempower, instead of selling it?

In comparison, many proponents of the EiP (e.g. @kevinwong, @trafalgar and myself) still have the majority of their stake on Steem. Which is why we want to make sure Steem becomes a success, as we've got a pretty big incentive.

Now, I'm obviously not arguing that risks are involved with these proposed changes and that discussions are important, but just talking will result in no progress => and doing nothing will be the death of Steem.

The solution is to build a new paradigm that will fix the fundamental issues with Steem, being:

  1. Plutocracies have always failed social networks of any kind.
  2. Investors and content creators have contradicting interests that are not being served by the same system.

I get it, none of us have had the vision to come up with a solution just yet, so meanwhile might as well try some quick fixes and see how things go. I'm on board with all of that. I'm just raising some questions that may help optimise said quick fixes.

Also, some clarifications are in order. In the past I've invested heavily in Steem. When everyone was exclaiming doom and gloom when Dan left, I bought hundreds of thousands of Steem. I always had significant liquid Steem, but for the longest time, I was also significantly powered up across 6 or 7 accounts (since you're snooping around, feel free to check @liberosist, @meerkat, @dunia etc.). While powered up, I used all of my stake to exclusively support curation initiatives. I can understand why you'd think I have an ulterior motive, and I won't call it libellous, but I've contributed significantly to Steem through 2016-17 and I'd urge you to believe me that I still have the best interests of Steem at heart.

To be transparent, I powered down and sold a lot during the bull run, but you should not assume that just because I am not powered up that I'm not a Steem investor. I started buying back once the market had settled, and I bought some more Steem as recently as a week ago. All I'll say is, I'm a significant investor of Steem, and the best interests of Steem are in my best interest. But the price isn't my concern here. A bull market will come around, and STEEM will gain significantly as usual. Doesn't really mean Steem will become a better social network.

PS: I'll absolutely power up when the two fundamental issues I list above are even partially solved.

Downvoted because the top voted comment does not have much to do with the presented arguments and instead dismisses what was written based on "not enough stake".

While nobody really knows the outcome of these changes, I remain optimistic that these changes may move things towards positive results.

Changing rewards split to 50/50 is the only thing that makes sense to me. The other two components of the proposal I don't care much about, however, if the whole bundle is what it takes to get 50/50 I am for it.

Honestly, I am surprised that you are opposing 50/50 rewards split. I might be completely wrong, but here is my attempt to convince you that 50/50 split will have positive benefits for participants:

  1. Organic voting vs vote selling. Over the years more and more SP moved to passive income that is generated via SP delegation and vote selling. Reasons are obvious, it is a choice of 25% in curation rewards or 4 times more than that. At the current state, SP will keep moving to delegating and selling rather than upvoting authors for free. 50% curation rewards will give organic voters a fair chance to compete. Moreover, it will bring some of the SP back to organic voting.

  2. Rewarding those who have been rewarding/supporting authors. While some chose to delegate and sell SP, some remained true to organic upvoting and continued supporting authors. They didn't have to. They could have gone a full return passive income route. But they didn't, because they have a long term vision. So these stakeholders, I think, deserve better rewards on their curation. I will be glad to have less rewards on my posts if that means those who have been upvoting my content get more. More SP these folks have better for authors.

  3. Will encourage authors to engage with their audience. Authors are also upvoters. They don't have to solely rely on the author rewards. They can meaningfully reward those who engage with their content, and receive 50% curation rewards. Many authors do anyway. While they may be losing on authors percentage, they will be gaining in curation rewards.
    This also will encourage authors to power up more SP.

  4. Will make Steem a more attractive investment. At its current state those who buy Steem, do so for speculative reasons, passive income via delegation or vote selling. Rarely you would see people buying Steem and powering up because they want to reward authors. 50% curation rewards will make Steem more attractive to buy and power up. If more people are interested to buy Steem, prices will go up. Authors will see increasing value in their rewards, stakeholders will see the value of their stakes go up.

  5. Majority of social media users are content consumers, not content creators. The way Steem works is, it heavily rewards authors and not content consumers. This in return turns content consumers into creating content to take some share of rewards. 50% curation rewards will let content consumers just focus on what they do best, consume content. This will encourage content consumers to buy more Steem and be rewarded for consuming content. I know quality authors put in a lot of time and effort to create content. But those who buy Steem, also put in a lot of time and effort to earn their money elsewhere and then invest in Steem/SP. Steem needs more people interested in putting their hard earned money into Steem.

In conclusion, I think 50/50 split will encourage more organic voting, divert SP from vote selling back to upvoting for free, improve content discovery, increase interest in buying Steem, increase interest for authors to power up instead of selling earnings, and give Steem a better chance to succeed.

I've explained this in the comment linked above. You're ignoring that bidbots will also see 50% curation rewards. Similarly, to self-voters, it doesn't matter, they'll get a 100% split every time anyway. That said, I agree it'll likely see a marginal gain with greater incentive for curation. It's just not enough - Steem has fundamental misalignments that need to be addressed. I'm sure I told you about this first in 2017, and through the years, and nothing has changed since; so I'll avoid beating the dead horse.

After reading your linked comment, I am not sure now. Your assessment and arguments are convincing. That also makes me think what will all the projects that are receiving @misterdelegation delegation do with extra curation rewards.

They will get 100% richer than they were before.

The issue of simulations has been brought up but there's really no good way to actually do so. They will run the numbers based on the current distribution but this tells us nothing about how the equilibrium shifts. Simulation in SMT doesn't give the same effect either. Maybe when AI is good enough we can have them mimic us :P

What have been discussed is exactly what you have noted, relative incentives. Many people have varying opinions on what is too much or too little but this particular configuration already appears to have growing support.

Sounds like witnesses have to much power and all users should be able to vote on any changes. Minnow to a whale..... Also we shouldn't have a downvote button. Just a flag but if you flag someone it should only be because of copyright issues. If your flag on someone isn't warranted then you lose your MANA for a month and cant transact nothing......

How the system works is that you elect representatives, i.e. witnesses. You're right that the problem is it's a plutocracy rather than a democracy. Only the very richest will get an influential vote, and most of them will vote for what they think will make themselves richer rather than a better social network. It's plain human nature.

@liberosist Yeah, that's sad. Not over @contestkings and our discord. Were building a community together for all to prosper thru contests and giveaways. We have one requirement to get on the daily whitelist for contests, No upvote, No Resteem, No Follow.... Thanks for the post, I appreciate logical thinking and opinions.....

Sounds like witnesses have to much power and all users should be able to vote on any changes.

You can vote on witnesses that represent your interest.

Yup, I'm thinking of taking down the two I voted for because they are getting to cool for their britches..

@yabapmatt and @aggroed are doing everything in their power to make Steem a success because they have got so much to lose if Steem would go belly up; removing their votes isn't making any sense TBH.

I never said that they are not, It's more important to have Witnesses striving for Steem than personal project's. If i ever tried or wanted to be one, that would be my only focus. My opinion, that's part of the problem.

They are striving for Steem.

Looks like we also need a downvote Witness button as well......

I'm not a Witness nor am i trying to be and i only deal with pumpkins at Halloween time....

I agreee that for now 50/50 is too soon for the STEEM blockchain. If you look at @penguinpablo stats You will see that Steemian voters OUTNUMBER content creators.... so to give content creators a 33% "hair cut" makes no sense right now... maybe maybe a 60/40 but 50/50 I don't think makes sense if you will lose content creators and give UPVOTERS less to up vote.

And for FREE FLAGGING, Bernie already does a great Job of it without the FREE, JUST ask Kawaiicrush or lyndsayBowes or anyone else who has watched the BOT WARS between various Steemians......oh boy!

I say NOT TO THESE CHANGES but I am not a witness.... maybe WE need a Witness POLL to see where they Stand and WE STEEMIANS can then up vote them in or out as witness if they support this madness about to be unleashed!!

I agree it's not the right approach. Overall, these changes probably make Steem slightly better, so I do not oppose them. But I also think they are not well considered and that most of them probably will not solve any problems. Implementing a basket of new things without simulating them is purely reckless.

EDIT: After further consideration, I've turned against this set of proposals also. They are a bad idea.

I think inaction would be worse. I am optimistic about 50/50 rewards split. Not sure about the impact of the other 2 components, but I think 50/50 rewards split has a potential of making Steem a more attractive asset to invest in.

 3 years ago  Reveal Comment

"...any change is good at this point..."

That's clearly impossibly false.

Why doesn't dTube work for folks on laptops? I can't view videos on dTube, and I've never received a response from you in most of a year's seeking answers here, on Discord, and email. FFS, at least acknowledge that there's an issue or that I've broached it with you.

Thanks for responding. Sorry for my tone, but I've been unable - and continue to be unable - to view dTube video for a long time, and have not previously received response.

I cannot today view dTube video.

Thank you.

That's an interesting thought. I agree that it'll cause a short-term disruption, but then soon enough people will find the best ways to optimize their earnings.

I'm fully in favour of making iterative changes, but iteration can only happen with... iterating. That requires making one change at a time, testing against a control, and studying the results of said change. Since HF16, we have had all kinds of things bundled in, and any empirical evidence to suggests what works becomes obfuscated.

Things are so broken here that I'm almost going to buy the argument that even a potentially bad change is good at this point. Almost...

As I've mentioned in a different comment, eSports games work because exploits are fixed, excessive abusers are banned, and countermeasures are implemented; the key being all of the above is done in a very timely manner. Steem will need a new paradigm needs to be implemented if it wants to achieve what eSports and pay-to-win games have. It's pointless if each change takes 6 months to implement.

Agree. These changes will not improve Steem.

Posted using Partiko Android

That remains to be seen. What I do know is there are significant doubts and risks associated at this stage.

Low ranked witness or not, your opinion is just as valid as @aggroed or @timcliff. We're all in this because we want to see Steem as a whole succeed.

What I would encourage you to do in the case of HF21 would be to propose changes that you think would be a better use of the HF. You probably did that in the post you linked; if you did, pardon my effrontery for not reading it. But you might be able to rally others to your ideals of you provide an alternative solution. Even if it's an idea you think is terrible, someone else might find value in it. They could help you refine it into something better. And I, for one, would gladly support a witness who comes up with solutions or at least tries to find them. Even if they fail to find a solution, a little effort in the right direction is better than none at all.

Posted using Partiko Android

My opinion does not really matter, while theirs does. In voting for hardforks, each top 20 witness gets 1 vote, we get pretty much nothing.

That said, if I can reach out to even a few people, it's a net positive.

I have spent hours in the last week brainstorming, offering alternative ideas, and more. Please see my blog and comments for more. I feel I have done everything that a low ranked witness could. Ultimately, I'm just a small pawn in the game, and I know my thoughts will likely never be considered seriously by anyone important.

I disagree in the assessment that your opinion does not matter. Just because you don't get to vote doesn't mean that the witnesses that can don't listen to those around them. I know quite a few top 20 witnesses, and they do listen to feedback, even that of a minnow like myself. And if you feel strongly enough about it, I bet they'd work with you to evaluate your solutions you've blogged.

Posted using Partiko Android

Look at my witness approval, I'm at #85. That quite literally means I'm unpopular and my opinion isn't valued anywhere near as much as a top witness. Just stating facts here! Nevertheless, I'm grateful for anyone voting for me, even if it's relatively few, and for them, I'll continue to state my opinion sincerely.

It doesn't mean that your unpopular it means that you haven't been voted for. I've been here for two years on the blockchain and this is the first time that i've seen your name anywhere. Now i'm not a big account but I've been involved in lots of different places here and have come across all of the top 20 and a lot more and seeing the work they are involved with. That is how I determine my witness votes from people that seem to be working towards the same idea that I have in my head.

This is why dPOS Systems need effective democratization of the social layer.

Decentralization is binary, they have technically no central authority, but they are highly clustered to their permission set of BPs. They only can play their advantage, when you distribute the power on the social layer. If not, they are nothing but a corporate-consortium and corporate consortia are BS and not Crypto. But once a System like this is running, you can't change a damn thing, because you have to convince those in power to give up their power. This is why Dan left Steem. Even if you are one of the 21 super representatives, you have to convince the majority.

First, build, then test, then run. Since Steemit was an experiment and one can simply move on it does not matter...but to us, if we want a future, it should matter.

Decentralised systems can work for corporations, economies, and platforms. However, once you're talking about a social structure, it just doesn't scale. Social systems need to account for human behaviour. Maybe decades in the future, an algorithm would be able to understand humanity, who knows. But in the here and now, I agree with you that the social stuff needd effective democractization away from plutocratic rule.

"...once you're talking about a social structure, it just doesn't scale."

I don't agree. Everyone has certain identical needs. For such needs, completely decentralized mechanisms can even produce consensus. The problem comes when less ubiquitous purposes are put forward, because not everyone has identical needs across all matters.

What has been proven to fail 100% of the time is any form of centralized control. Over time, every empire falls. Freedom - decentralization in other words - is never completely eradicated, but even under the most oppressive tyrannies, remains in cracks and hidden in shadows, to burst forth once the despotism finally so enrages enough people that it is overcome, to burgeon once again.

Until some new centralization of wealth and power is implemented over people.

Freedom will eventually vanquish all oppression. Decentralization will eventually completely eradicate centralization, and we'll be free. It's the only form of society that cannot be defeated, and we're just waiting for the technology that enables it to be developed to maturity.

Decentralized systems have proven to be the most centralized, time and time again. STEEM's GINI coefficient is far worse than even the most centralised dictatorship in the world. (I.e. Brunei)

ja, most traits are maybe not equally but normally distributed. The need for communication, entertainment, even our taste, and our ethical views. Also, skills are normally distributed. But information follows power laws. Followers, Money, Influence, ... there is only one pewdipie with 96 Million followers. Not 1000, not 100, only one. And it does not matter if you start with equal conditions. After n rounds there will be one node holding most of the resources.

It's maybe not impossible to scale. Make stake private, make voting contribution anonymous and introduce unique user identification so that self voting is impossible, even with multiple accounts. Stake should only give you the possibility to earn return. Its Proof-of-stake, not proof-of-status, and proof-of-brain and not proof-of-egoism.

This proposal have been in history of tax reform. Remember, how tax is created. Government say before, that the tax is made to punish the only the rich or whales.

A reason why this minnow or the poor and the middle class agree for tax reform and the law is made.

Now look what happened, the minnow or the poor and middle class pays more tax and the rich don't.

HF21 is somehow in my own understanding is exactly alike how tax is made. Correct me if I am wrong.

You're actually spot on. A convergent linear curve does mean minnows have to pay a tax that the rich don't.

I note we disagree on the EIP, but we have honest reasons for that. The downvote proposal is cancer. I'll let the votes and comments of others on this very comment make my argument for me. I'm confident they will prove beyond a rational doubt that free downvotes is a bad idea, and may be the straw that breaks the camel's back regarding Steem value, user retention, and censorship resistance.


Remember, one abusive whale can effective now vindictively kill off hundreds good content creators at zero cost.

Again, share the same thoughts.

And my witness vote since i forgot it last time.

Hey, @liberosist.

Based on atleast one comment, I guess I'm not the only one wondering who HF 21 became more than just the blocktrades worker proposal. I did take a look at github, and while I really have no idea what I'm looking at, I did find some mention of a convergent linear curve being worked on. So, there's that.

Thanks for speaking up. I share your concerns, particularly around the downvote pool. The potential for bad behavior stemming from there is substantial.

⬆️What this smart individual said.

Thank you for standing up for what is right for the Steem network! <3

While reading your post, I got an idea how to fight bid bot abuse: add a small "fee" for delegations, devalue delegations a bit, while creating an incentive for manual curation:

I'll tell you a secret. I deserve a tiny bit of the blame for the SP delegation feature. Back in 2016, a few of us started Project Curie. There were whales who would support by trailing our votes. Since there was no delegation feature, we worked out a deal where a script would track how much rewards Curie's votes were generating for their account, and they'll donate some or all of it. The SP delegation feature was designed as a response to make this easier. (It also had other drivers, such as making account creation with delegation possible, of course.)

Anyway, while I agree with your idea, ultimately that's simply what will happen if there was a delegation fee - people will go back to how it was before the feature existed. They would just follow the voting trail normally, and have a script do the calculations for the rewards generated etc.

very interesting and thx for the feedback, simulating delegation with a script and sharing rewards sounds interesting but also a bit complicated and perhaps would stop some users from using bid bots and just curate themselves. Because Steemit is a very complex and dynamic system, I think we should try out some promising proposals, test them for a predefined time (for example a week or one month), and if the results are good, keep the change or otherwise roll back the change.

But EIP isn't HF21(unless I'm mistaken), thats DAO(the steem proposal worker thing by @blocktrades).

Seems like all of it is being bundled together for HF21. Point taken though, as they have not officially annoucned it. Regardless, my opinion of all of the above is stated above.

I really hope they don't bundle it all together. More frequent, but smaller HFs would be nicer so we can stop the debacle that was HF20.

Unfortunately, this hasn't been the case since HF16. Every hardfork since then has come with all kinds of things bundled together. They see each hardfork as a risk, but of course, paradoxically, bundling all kinds of things into hardforks actually increases the risk.

All I'll say is - a) SPS code was completed a long time ago, and b) they are not talking about releasing SPS hardfork, but rather debating EIP and c) if you see their GitHub activity, they are coding in the EIP changes for the next hardfork. Make of that what you will.

I liked your post and the numerous discussions here about HF21, this deserve an upvote and a resteem from a plankton.

I’d like to see a consensus to try this new system and if the abuse becomes rampant have a rollback plan that can be implemented in under a week. I think we need to keep trying to fight bad curation and this just might be a step in the right direction. I completely get your concerns though.

Posted using Partiko iOS

Congratulations @liberosist!
Your post was mentioned in the Steem Hit Parade in the following category:

  • Comments - Ranked 2 with 69 comments
 3 years ago  Reveal Comment