Steem Hardfork 20: Thoughts on Velocity
After 14 months, we finally have a new hardfork to consider for approval.
My feelings about this new hardfork are currently mixed. Unfortunately – in typical Steemit, Inc. fashion – this is another let’s-cram-it-all-in-one-proposal hardfork. Before I even begin, I’d just like to reiterate my desire to have more frequent proposals and less changes to consider for each fork instead of the exact opposite, where we have to either accept undesirable changes or have no changes at all.
With that out of the way, I see some good and necessary changes being proposed and I see some things that appear to be unnecessary and overcomplicated.
What I like about Hardfork 20
Removal of the minimum SP power down restriction – If the SP is yours, you should be able to withdraw it. If there were concerns about exploiting account creation, then the fix is to mitigate account creation exploits, not to prevent all users from withdrawing their tokens. This one is a no-brainer for me.
Reducing the curation window time and self-voting rewards during that time (reverse-auction) – This is something that I have supported since the spring of 2017. I’m glad to see it finally being proposed in a hard fork. I honestly have no idea why it took so long to make this change. A 30-minute “reverse-auction” window was unnecessarily long and has done very little to nothing at all to “combat bots.” Self-voting within that window and being able to maximize your rewards was never ideal and didn’t make much sense to me. And anything that may increase the total pool of curation rewards is a great step. We definitely need to improve SP utility and SP returns for invested users.
28-day expiration of internal market orders – A good addition, in my opinion. I don’t think it’s a critical need, but if it helps to improve blockchain performance with insignificant or no negative consequences for token traders, then I’m for it.
Reward beneficiaries paid based on author settings – Again, this is a no-brainer and a good implementation. It keeps things simple and uniform, which should not be underestimated (time and again).
SBD print rate – This just makes sense to me. If our debt limit is 10%, then we shouldn’t be reducing the print rate of SBDs starting at 2% and increasing the printing of STEEM. In the few cases that I have observed, it’s negative STEEM price action that leads us closer to the debt limit, not the over-printing of SBDs.
I do, however, have a major issue with the SBD conversion function being taken off of the flagship site, steemit.com. This conversion function is the primary tool for users to combat the oversupply of SBDs. It isn’t an “advanced setting.” It’s a critical blockchain function. With prices creeping toward $1.00, it may be very necessary to use the conversion function to both reduce the downward pressure on SBD prices and to reduce the debt ratio by slightly increasing the STEEM supply at the same time. If having a pegged token is still desired, then we need the available tools to be...available.
What I don’t like about Hardfork 20
The new account creation functions and features.
I’ll keep my critique of this simple: It all seems to be an unnecessary and convoluted mess.
Back when the 2017 road map was released by Steemit, Inc., one of the concepts that they touted was K.I.S.S. – Keep It Simple, Stupid. It was a great idea that had a lot of support, as the blockchain already seemed pretty complex and easily confused new users. Making things simple for users, making the blockchain more efficient, keeping things as transparent as possible, and allowing for quick troubleshooting are all excellent ideals to strive for with tech and social media.
We’re not getting that here.
Instead, we’re being asked to accept a more complex system of account creation that will require new parameters from witnesses, including many of the same ones who are unable or unwilling to set existing parameters today. These parameters are supposed to be set for discount account creation tokens – the number created each day and the total supply. Then the new Resource Credits (or RCs, which are being proposed in this same hardfork to control bandwidth resources) will apparently be used to actually create the accounts.
Very few of the top-50 witnesses even bother setting bias and APR parameters, even when SBDs are off the intended peg. When it comes to block size and creation fees, most of them simply wait and copy whatever the first person does. If we’re hoping that new parameters will be set based on robust testing and discussion among witnesses and the community, then I’m afraid our hopes will be sorely misplaced.
Even if we were to assume that all of our top witnesses will be engaged and discussing these new parameters regularly, these account creation changes alone should be their own hardfork proposal. There’s no reason why they ought to be included with curation reward, power down, and internal market order changes. The account creation and RC system/market are significant alterations of Steem protocols and will likely have a large impact on the ecosystem. And as stated above, they will add a thicker layer of complexity to the system.
After testing on a testnet, it’s possible that it may turn out to be a better system than we currently have. But why it needs to be lumped in with other quick/easy changes that have more widespread support is mind-boggling.
All of that being said, I think the Resource Credit idea is worth testing. I just don’t think we need another 20 changes at once...again.
20-second comment limit reduced to 3 seconds – I’m not even sure why this was something that needed a fix. Are spammers not able to spam quickly enough with their spam bots? Or do we simply have world-record speed-readers and speed-typers out there who really want to leave engaging, meaningful comments every 20 seconds on different posts?
Upvote lockout period changed to cool-down – I don’t dislike this change. But I don’t necessarily like the time frame for the cool-down.
Very few posts receive votes (or any attention at all) after the first two or three days after publishing. This was the case before the 7-day payout was even implemented – users just didn’t engage much with older posts (mostly because users don’t engage much at all, but that’s another issue altogether). This was one of the main reasons why I wasn’t exactly enthusiastic about changing the payouts to seven days. I would have much rather had a 48 to 72-hour payout period with extended voting time given to the post based on votes received toward the end of the payout period.
With a seven-day payout and a lack of voting and engagement on older posts, I would rather see a cool-down period of 3.5 days. This would help avoid “abuses” in both directions. It would limit the amount of late upvotes and/or votes purchased via bid bots that are made with the intent to try to hide the rewarding of the post after it is long out of sight by the average user. And it would likewise mitigate any late-stage downvoting that may be meant to simply “punish” another user.
More than that, having the shorter initial window would help with the notion of discovery and creating viral or trending content, which was the original idea behind the reward protocols in the first place, if I’m not mistaken.
So I like the idea of the cool-down as opposed to the lockout. I’m just not a fan of the current proposed implementation of it.
The rest of the proposed changes
I really don’t have an opinion on most of what was not mentioned here, which means I don’t really have a problem with any of it at this time. It’s all pretty mundane.