You are viewing a single comment's thread from:

RE: On Trains and Onboarding

in #hive8 months ago (edited)

Just a few aspects which could prevent the success of HIVE in my opinion:

  • I think one of the HIVE problems are trails and auto votes.
    A few powerful 'trail leaders' choose a few selected 'good users' who are getting upvoted again and again, while the majority of new users simply remains undiscovered and rarely receives any upvotes.
    In my opinion upvoting should be more 'decentralized' in a sense that every user has to make their upvotes themselves and completely manually only, because automated upvotes means to comfortably select a few 'good' users and friends again and just upvote their posts without actually reading them.
    We all simply should do the hard task and read through posts manually day after day, while not selecting them by the criteria if written by 'good' and 'bad' users but by their quality (to prevent that 'good' users are getting lazy and 'bad' users getting motivated to become 'good' users)!

  • Another improvement would be to allow downvotes only in case that their original purpose is fulfilled: to fight spam and plagiarism.
    I actually also get the sense of downvotes which serve the purpose to 'redistribute' the rewards pool by downvoting 'over rewarded posts' or users who publish lots of low quality posts.
    What is really bad for HIVE are downvotes given because of different (political) opinions or just simply personal antipathy. Especially counterproductive are automated downvotes where the attackers don't even read the content of their 'victims' but just flag whatever they write.
    I think in cases like this the 'victims' should be able to call a committee of highly respected users (with a rather big stake of delegated HP) whose task would be to counter unjustified downvotes. (Another option would be to try to solve that problem by AI).

  • Concerning your Hivewatchers I would wish to see more interaction/explanations in case you downvote new, unexperienced users. First tell them what their mistake is, and only next time downvote them in case they didn't react to your explnations/warnings. Also better answer to their questions on the HIVE blcokchain (not only in Discord).
    I should add, that you are doing an important job, though!

  • Furthermore, in my opinion the HIVE government system is not 'permeable' enough. In a vivid democracy politicians have to change from time to time to prevent the system to 'freeze'. However, in HIVE I more or less see the same user names in top witness positions since I started to use STEEM in 2016.
    The reason is the concentration of power in a few hands, which partly stems from early mining from the early STEEM days and partly from the extensive use of bid bots. Even if these times are over since quite a while, they still affect the current state of the HIVE eco system.
    In addition most of these old witnesses are connected to each other and support each other. That of course doesn't make them 'bad people', but nevertheless is rather bad for a real decentralization of HIVE.

  • I personally think that 13 weeks of unstaking one's HP is just too long (even if I understand the reasoning behind it). You simply cannot use your HIVE in case you would need them, and most investors simply don't accept such a long unstaking periond. I think a time span between four and six weeks would be more realistic.

Sort:  

Most of your points are about increasing communication. A lot of this can be achieved through frontends -- a custom field to fill out when upvoting and downvoting which would be recorded along with the vote as part of custom json. It's possible. The trails could also be managed in the same sort of way but all they'd say is 'trail vote' or along those lines (where the trail is controlled through a service and not a personal bot). One issue is that voting would be more 'expensive' simply because instead of one, two transactions would need to be sent. This wouldn't affect most people however.

In regards to the same witnesses being there it's largely to do with the fact that witnesses have to work together, understand the protocol code, and it's generally not inexpensive in relation to resources. Around hardforks, for example, I have at least 4-5 nodes on different servers running. Every hardfork we end up seeing the same people engaged from both the consensus and backup witnesses. In the pre-Hive days, access to witness communication resources wasn't accessible to all, it was tiered. Now it's all easy access from various platforms to the same chat. Hive completely changed all of that.

I agree in regards to the power down/unstaking, it would be a good idea to start projections of what lowering it could look like.

Some valid points.

When I see discussing Elon Musk on X how to fight spam, and if - as a consequence - he should take a member fee of every user who wants to write there, it shows me how important platforms like HIVE are which already offer the solution to problems like that (and others).
But anyhow, the hurdle for many people to join and use HIVE still seems to be too high ...

 5 months ago Reveal Comment