My opinion on the upcoming v17 hardfork.


For reference on the V17 update, please read this first!

    I have had a few people ask me already if I will be supporting Hardfork 17 as the date draws near. Many other witnesses have come out in with their own thoughts, and each is worth a read to start to understand the current community impression of the fork. Here's a few I chose that I think are worth reading:

@klye -- /@klye/lost-top-19-spot-by-showing-support-for-hf17-and-falling-victim-to-old-boy-mentality-fml
@bhuz -- /@bhuz/my-take-on-hardfork-17
@abit -- /@abit/steem-0-17-0-released-but-i-won-t-upgrade-as-is
@dragosroua -- /@dragosroua/i-didn-t-sign-up-for-this-shutting-down-my-witness-and-seed-nodes
@pharesim -- /@pharesim/rejecting-hf17-and-a-few-thoughts
@chitty -- /@chitty/to-my-witness-supporters-lets-vote-on-hardfork-17


So here's my own thoughts on each of the changes.

  1. Remove Posting Rate Limit.
    • This is a good change, as it is something easily counteracted by sybils anyway. My point of view from an abuse side, being a part of @steemcleaners, is that we play wack-a-mole with spammers who don't care about this limit anyway. A post limit that makes it complex for a normal user just is not worth it. After this change @steemcleaners will still be playing wack-a-mole with sybils.
  2. The comment depth limit has been increased to 255.
    • This is useful, and it can be up to the interface how to determine how replies are shown.
  3. Comments can now be permanently edited.
    • It says comments, but I hope this applies to all content including posts. This is absolutely necessary, as there is no other blogging platform or website that says "oops! your own content that you wrote and own can't be edited, sorrryyyyy!". Very happy with this.
  4. Comments are now paid out independent of their discussion.
    • This is alright. Attaching the comment payout timer to the original post doesn't make much sense. We should be able to a) comment on an old post, and b) be able to be rewarded for commenting on an old post. Consider an old blog post that suddenly goes viral: We want the author to be able to fix spelling mistakes, and we want to be able to have comment interaction. Comment interaction includes new comments being applicable to be rewarded.
  5. All comments are paid out 7 days after creation and there is no longer a second payout window.
    • I don't quite like the 7 days for the reward. It's simply too long for the average user to understand what is happening: People have short attention spans, and the volatility of reward is incredibly confusing to new users (Post valuation changing over time -- although one remedy to this solution is to only show users received RSHARES, and not the value in dollars. This would be a huge flip in UI and marketing however, as people are currently used to seeing $$$ values.). However, it is also incredibly hard to reconcile the "ability for old posts to go viral and recieve engagement", as I just mentioned, as well as the original post being rewarded. From a consensus point of view, there is no way to be able to continuously vote on content (this would balloon the amount of data to keep track of, forever). We need to think of a better way to, a) reward users promptly, and b) reward users for old posts that either get refined or go viral. Consider the welcome guide to Steemit: further rewards to the writer would be nice, as would the ability to modify and edit as steemit changes, and finally comment on the post like any other blog post. However, it is hard to reconcile a) and b) together, and I unfortunately do not have a solution.
  6. There is now a comment reward fund separate from posts.
    • I don't particularly like this change. This is against the idea of simplicity, this creates a new market of trying to balance rewards (curation, and author) from two different pools with different reward payouts. This is immensely complex for an investor to maximize returns, and will likely cause a teeter-totter of people trying to game rewards from one pool or the other.
  7. All payouts now look at the prior 30 days of payouts to determine the share of the reward rather than the current pending rshares.
    • I understand the technical reason for this change, and I think it will be fine when we reach a steady state.
  8. Reward Balance
    • What is this? I don't understand the reason for this change, nor any communication about this change...
  9. Comment Reward Beneficiaries
    • Nice! I like this a lot, it allows people to reward other people for their own posts, and allows interfaces to claim a small fee for posts using the interface in order to pay their costs. This will be especially great for Busy, and web apps.
  10. Delegated Steem Power
    • This is cool, and will allow blockchain level support for people wishing to let others use their voting power. Trusted curators can be lent the power to vote for content, people who leave for vacation can let their friends vote for them, etc. All this without the requirement of running a bot curation trail (not accessible to the average user!). I hope people who currently run curation trails will consider using this option instead of trails, as the "vote count" is completely insane right now -- a trail could add 100+ "votes", when in reality it was only one person, the curator, who read and enjoyed the content. If people choose this option, we will see a more accurate vote representation. However, if the delegator does not earn a curation reward, we will likely still see curation bots, unfortunately.
    • If you are a follower of my curation trail on streemian.com, and are reading this post -- let me know in the comments what your thoughts are on this: Would you still give me your power to vote if you did not earn rewards? Would you delegate me your power if I payed you a cut of the curation rewards I recieve? Or would you still use streemian as-is?
  11. Accounts can be created with a smaller fee and an initial Steem Power delegation.
    • This is interesting, but I question if this will allow name squatting for people whom have plenty of SP to delegate.
  12. PoW is being removed.
    • It's clear that PoW is broken in its current form. It is also clear that PoW does not actually help secure the network (As it is only one block in a 21 block round). We have never been a "hybrid" PoW/DPoS to be honest, as mining was never a significant amount of the block reward. HOWEVER, one thing people forget about mining is the ability to create new accounts anonymously. We have tools like anonsteem to do this, but this requires trust in an external authority that you pay to recieve an anonymous account. I wonder if we could think of a blockchain level replacement for this, to satisfy me with the removal of PoW.
  13. NTP is disabled by default
    • Sure, more technical improvements. Moving to ntpd is good.

Phew, long read. Still with me? I have more to talk about.


On Witness Politics.

    As a relatively new witness, I am not a whale. I did not get my witness spot until just before the hardfork that cut witness pay to a fraction of the amount. Not being a whale in the witness list means that I do not have the ability to do a "tit for tat" to keep my witness spot -- we regularly see whale witnesses voting for each other in a game theory way (if they vote for me, I vote for them), but this ability does not extend to those of us in the witness list that are not big whales. Due to this, I consider my witness spot more of an elected position, and I have a responsibility to listen to the people who do vote for me as a witness. Although this normally would be a good thing, at the present moment this is actually also concerning to me.

  • If I do support HF17: I may be voted out of the witness list by members of the community who are not supporting it (e.g. abit).
  • If I do not support HF17: I may be voted out of the witness list by the employees of Steemit, Inc. whom vote for me (e.g. ned, val, roadscape).

    Personally, I think we need a better solution for determining the progression of hardforks. Regularly mixing hardware changes (e.g. NTP tweak) with community changes (e.g. Comment reward pool) is not a good thing, and should be separate. We also might need something like community voting -- which could be implemented by witness voting (elected representatives) -- however there represents a conflict of interest if we do this while Steemit employees vote (and I would argue that dan/dantheman perhaps should also choose not to vote if Steemit employees abstain).

    So, I currently do not feel I have a good enough grasp on the situation to take a hard stance at this time, and thus I look forward to continuing to read comments and discussion to help me make up my mind -- a decision that will be "on behalf of my constituents" . If you support me as a witness and have strong opinions about the changes, please comment below!

H2
H3
H4
3 columns
2 columns
1 column
44 Comments