You are viewing a single comment's thread from:

RE: Hive core developper meeting #10

in #dev4 years ago

Sorry for not replying sooner, I'm quite swamped with hf24 things.

I really appreciate these and just got done listening.

Thanks ! I'm always a bit like "is this really worth it ? It's pretty much a podcast where blocktrades and I talk about stuff who cares ?"

  1. Great !
  2. I feel like this could be solved by writing it in the json_metadata of posts And then we would index that in hivemind. But first we need to have all front ends add a field where users can select a language for their posts. But ultimately you mostly need to convince @good-karma and @blocktrades since you guys own the major front ends. Then you could open an issue on https://gitlab.syncad.com/hive/hivemind to request the feature. I also feel like that would be a good candidate for a machine learning project where you just add a language detector that works with hivemind so that the burden is exclusively backend and not on the front ends / users
    3.4.: Not sure where it's at, but yeah I agree that we should work on the first protocol upgrade for communities, good thing is that it's layer two so it could happen in parallel to the regular hard fork development.
  3. Nice !
  4. As we said, I'm in for a complete separation of the reward pools from the base token. Blocktrades doesn't seem to be as drastic as I am. Other than that there are probably more experiments that we could try running

Mhh I agree with you on the paywall, but it's a hard problem to solve while being decentralized. A first centralized version could be done where the content is encrypted by the hive-patreon front end and only decrypted by the front end if the user is set to send recurrent payments.

But that's a centralized solution and it's not great. I'll have to brainstorm this, I think it's hard but not impossible.

Sort:  

5. Patreon and pay wall

I think having a decentralized option to begin with is great however giving our users a Choice to add encrypted posts to a centralized repository is viable... maybe at least a 3rd party or allowing the user to choose between a couple options? I dont think platforms like ourselves want to be responsible for storing sensitive material and sometimes illegal content.
5b. The big issue is how each platform gives the reader access to the encrypted content perhaps. We check if pay wall requirements are met then become the arbiter if the reader is allowed access? Or does the creator somehow directly impact the access? Perhaps the creator decides to trust peakd for example for us to verify who has access to content and then gives us the ability to decrypt for that user. That is not the perfect system of course but i dont know of better ones.

  1. We did come up with some api related errors I think asgarth reported a few days ago and hivemind was being redone anyway to fix them.
  2. Yeah we can add the json metadata for language easily enough and we can even identify what language they are writing in and automate the setting in our advanced section (allowing them to change if they want i guess ) so users rarely if ever have to actually select it themselves (unless they want to)
    2b. The big part is just indexing in hivemind... and maybe that just requires @blocktrades and @good-karma to want the language feature and I'm sure we are willing to test it out early on letting users know that not all content on the chain will be shown when they do a language based feed for example.

Honestly the indexing part shouldn't be too hard to do we already parse the json metadata, it's just a matter of adding a column.