Discussion about antispam mechanisms

I feel like we need a discussion about Duniter antispam mechanisms and the price we put on it. It is linked to the first discussion Comment partager équitablement cette ressource commune qu'est la blockchain Ğ1?, but I think we are better prepared for it now.

My feeling is that the current approach is too much focused on the rare hypothetical case of spammers rather than the everyday use of people. Also, there are different kind of risks:

  • filling a bloc capability → prevent other from adding transaction to it
  • increasing blockchain size → force smith to have larger disks
  • increase chain storage → increase ram use / cost of extrinsics (??)

Each one deserve a different mechanism.

The risk of migrating Ğ1 without sufficient explanation is to harm the trust of the community, and I consider this more harmful than a spammer using tons of Ğ1 to try block the network for an hour or so.

A list of linked topics to provide food for thoughts :smiley: