Thinking about Chris Aldrich and the discussion around Agora that I was also having with Vera; I realized I wanted a place to lay out attributes of how I would want federated notes to work.

give me markdown or give me death

I think there are enough tools to convert from other formats to markdown that this can work for most people. I like that Agora takes a strong stance on this.

multiple typed relationships

You know how certain issue / ticket tracking software lets you specify the type of link between two issues or tickets? Jira has four types, one that I use in my life uses more – “is caused by”/”causes”, “is fixed by”/”fixes”, etc. I want flexible types of relations between notes along these lines. “This reminds me vaguely of this” separately from “I found this when I was looking into this” separately from “I think this is the same thing as in this other domain over here”.

I want this because in my notes, I find myself writing over and over “seems similar to [blah](#blah)” just to get a link in the note that will then cause the connection to be displayed as a backlink as well.

Probably has to be free text (maybe taggable?), displayed distinctly from notes, but ultimately…

relations between notes can be made into notes

Okay, so this may seem unnecessarily complicated, I guess, but… The important thing about notes being a graph for these purposes is that you can add on new nodes and edges connected to other people’s chunks of the graph. Sometimes you’re going to add a node where what you really want to respond to is the idea of (or the text associated with) someone’s edge. An example:

If I have a node with a quote from Levin’s chunks of Anna Karenina, and I make an edge to my node on “VC Twitter” saying that I think there’s something related about their attitudes… Vera might come along and create a note pointing to that edge elaborating that a lack of self-awareness that what one is doing is “slumming it” does not actually change the dynamic of the pretension, and she might further link from that note to… I don’t know, some other node which constitutes a clever observance1 about some other instance of the pattern.

no automation of equality

(thinking about agora and this as well)

If I write something about, say, the Seattle accent, and someone comes along and creates an edge saying it’s “the same as” their note about West Coast accents (by which people invariably mean California), I don’t agree with that! This then means that there has to be some thought put into the UI about letting an author privilege certain edges other people have applied, while still allowing discovery of that person’s wrong opinions about accents.

This does not change at all if it’s two people using the same term to denote different concepts, which is why I cannot accept this aspect of Agora’s approach. I recognize it comes with a lot of benefits, but I refuse to use a system that demands my note identifiers be socially negotiated. (That wording sounds a bit more dramatic than I mean)

paragraph based, not an outliner

This is both because I use paragraphs for everything and because the flexibility that bullet-specific granular links allow doesn’t necessarily work nicely in multiplayer. A “follows” relation might be nice in order to thread together short notes for roughly the same effect.

  1. Vera may not hold this opinion at all at all about Levin from Anna Karenina, I just made it up for the example, sorry. Actually, for all I know, she may think highly of VC Twitter. But it’s late, I can edit this another day.