I will keep this short.
For reference, you can find the HF 17 Release Notes here
What I do not like
All comments are paid out 7 days after creation and there is no longer a second payout window.
99% of votes are cast in the first 7 days after creation. This elimates the need for a second payout to accumulate value to a post and simplifies logic significantly.
First, there was basically no community discussion about this change. The dev team just put it into HF17 and did not even care to reply to those witnesses that raised some criticism about the change well before the final release
Second, it is pretty obvious how the data presented is extremely biased depending on the current steemit UI and the underlying blockchain reward mechanism.
You can't really expect to see more than a few rare votes on contents older than a couple days, just because the UI is focused on showing new, recent and trending post on their first payout.
The 30days trending page is very rarely used (and now it is even removed from the site), and I think both because- the daily posts aren't so much that users end up seeing a post for the first time only after the first payout (so due to a low number of users and even lower number of content creators) and
- there is really no incentive to vote during the second payout period since there is no curation reward
Third, Steem and Steemit are two different things.
I can understand (I would not be a fan tho) if Steemit.con wants to focus only on certain kind of contents that seem to be "valuable" only during a narrow period after their creation and then "lose" their value entirely after just a few days.
This doesn't mean that the Steem Blockchain should limit itself and other possible competitor UI and websites to focus on all those contents that really doesn't lose much of their value for quite a long long time. Just think about useful Q/A, guides, tutorials and such... why should they lose value and possible reward after just 1 week?
It is definitely pretty common to read and use even a years-old tutorial, they are still useful and very valuable.
This third point shows how even the current reward payout mechanism is not ideal. If a change needs to be done, imho, it should be towards an open system where users can upvote and rewards content creators no matter how old the content is, and possibly do so even more than once per post.
The fact is that the reward payout mechanism is/could be a crucial point of the entire system, it needs to be designed more openly as possible to make it works for different kind of content and scenarios... Back to the first point, it really needs much more community discussions and open eyes on what we want in the future... do not limit the entire blockchain because of how you intend to use it just for steemit
- Fourth, security reason
There is now a comment reward fund separate from posts.
We want discussion to be incentivized. Currently, comments are only claiming 2% of the content rewards and there are more comments than posts. Comments will receive 38% of the content reward fund. This change should incentivize commenting, engaging in, and curating discussions.
Both witnesses and the Community pointed out how this is not needed, at least not now.
I am against directing a fixed percentage to comments, and even more if that fixed value is so drastically high compared to the current one. The fact that it is hard-coded and we would need to wait months for change it back to some more reasonable value, it just add another strong reason to be against it.
I think it is way better to have a single reward pool, leaving the distribution between posts and comments to be dynamically allocated with a new, better, much wanted, linear reward curve... A better UI for comments could help too.
Start KISS, and add some level of complexity only if real issues arise. Do not start it obscure and complex just because of theoretical issues that may never present itself or be somewhat relevant considering the whole system
What I would like but isn't strictly necessary now
- Changes to the Reward Curves
The only reason why I am saying that it isn't strictly necessary now is because of the no-whales-voting initiative.
If that initiative continues, the bad effects of the current n^2 curve will be (and already is) much smaller. If the dev team really wants more time to make it "right", we probably can wait until the next HF.
That being said, imho, it would be way better to postpone the HF17 to a few more days so the dev team can finally listen to the community and witnesses demand and include the changes needed for a simple linear reward curve. KISS.
Community to decide
If you support my view, you can also support my witness by voting for
bhuzIf the community decides to support HF17 as it is now instead, once the HF is passed I will still upgrade my witness node and keep supporting the network and the new changes as requested by the community itself through witness voting