It's unclear whether being able to insert a new identifier within an existing list of identifiers is a feature. That might just be a byproduct of needing to put

It's unclear whether being able to insert a new identifier within an existing list of identifiers is a feature. That might just be a byproduct of needing to put the new card into a related area. The Zettelkasten.de software appears to be pushing concepts that would have been impossible in Luhmann's system. Ultimately it seems that all these information systems are just different implementations of nodes, edges, and containers. Node: Do you want it to support formatted text and/or media? Do you want to be able to visually explore this node's edge relationships to other nodes? (Software often chooses one or the other; both would be nice.) Edge: Do you want directed edges? Do you want to support Graphs and not just Trees? Do you want your edges to have configurable meaning beyond "this relates to that"? Containers: Do you want to be able to collect multiple nodes together? Do you want a node to optionally be a container, meaning do you want to allow an edge to link a node to a container? Do you want to support non-hierarchical (Venn) containers, aka tags? Outliners, Todo list software are almost always directed trees instead of directed graphs. Even with mindmapping / bubble graph software - it's weird how many of them are just undirected trees instead of undirected graphs. Even orgmode is at best a directed tree with tags - it's not a DAG unless you use something like org-brain, which has its own limitations. Wikis usually don't have the visual exploration tools - imagine being able to zoom out of a wiki and just draw edges and containers to put the nodes (articles) in. Zettelkasten supported (somewhat) formatted text. Not really able to visually explore edge relationships without taking stuff apart. Undirected graph, generic edge meaning. Links probably often one-way and not self-healing (removing a card wouldn't remove links of all linking cards). Supported heirarchical containers, and a card could conceptually link to a container by having "2" link to "2a" which might include 2a1, 2a2, 2a3, etc. Supported tags, sort of, although there wasn't a way to search for all cards that had a particular tag. Ultimately you find that different thinking styles require different combinations of these concepts, so there isn't a one-size-fits-all, and there isn't "one system" that works like how the brain does, at least not until we have a system that allows us to easily mix these concepts together into a custom implementation.