Communication protocols that can be built on hinter-core

[bbenligiray]

It would be really simple to implement a rudimentary DM feature to the helper and I can imagine that being very handy for use cases where peers don't have a traditional DM channel (which we didn't come to experience yet) or that has become unavailable for some reason:

When Alice wants to chat with Bob, she can find the messages from Bob in peers/bob/incoming/.chat and she can find her messages to Bob in peers/bob/outgoing/.chat. However, she doesn't view or edit these files manually. Instead, she uses the helper, which parses them (possibly only the last N lines using tail) and interleaves them according to their UTC timestamps to display the correct order of messages, and she uses the helper to append a new message to peers/bob/outgoing/.chat as a new line in JSON Lines format.

This file-based design is intended to easily support alternative clients and AI integration. The base helper functionality should be fairly limited in comparison.

As a note, the peers/*/outgoing/.chat directories will not have an equivalent under entries/ (like regular report drafts) so the helper should not modify (i.e. delete) peers/*/outgoing/.chat while syncing reports.

[bbenligiray]

A similar kind of a functionality is a collaborative dictionary. Every peer keeps a dictionary in their entries (for example, in entries/.dictionary/) and have a peer group that they share it with, where the destination path is peers/*/outgoing/.dictionary/. To browser the dictionary, a peer runs a local client, which parses through all peers/*/incoming/.dictionary/ and displays all entries in merged form. This enables easy dissemination of information/opinion about people/organizations/events, etc.
This can be thought of a more realistic path for the regular user to provide content, given that they won't be expected to write up full-blown reports/articles.
One way this can differ from the traditional online collaborative dictionaries would be if authors were a bit more editorialized. For example, if publications maintained and shared dictionaries (rather than individuals) where entries are reviewed and verified by multiple people in a journalistic way, browsing a dictionary based out of 10 reliable sources (compared to thousands of anonymous Internet users) would be actually useful for quickly querying what people are about.