Org-mode alignment

[joshsh]

This has been investigated, but one of three things will need to happen:

• it becomes possible to run Brain-mode in Org-mode
• Brain-mode mimics Org-mode in terms of commands
• it is decided that Brain-mode and Org-mode are incompatible and will not be integrated

[JeffreyBenjaminBrown]

Reddit user attrigh suggests making smsn-mode able to read org-files and display them as if they were part of the graph. Once the data model has generalized to include files, such a file (and better yet, individual notes within it) could be linked to from the graph.

[joshsh]

I have been thinking the same thing, and reading up a bit more on Org-mode. I quite like its approach to folding, hiding, and moving nodes within a buffer. I once tried to embed brain-mode/smsn-mode in org-mode, but ran into some incompatibilities. We do now serialize atoms one-per-file, where each file contains the atom's properties and immediate children. I have been moving toward Markdown for these files, but am definitely interested in Org-mode as well.

[JonathanReeve]

FWIW, I've been daydreaming about how triple-based notetaking might be a useful addition to an existing tool like org-roam, since it's already most of the way there. I wrote about it here., and posted to the org-roam discord here. It'd be great to have a semantic triple based notetaking engine in emacs.

[joshsh]

I also considered triples for SmSn early on, but went with an unlabeled model for the sake of placing minimal constraints on the user; you don't need to worry about managing predicates, using the right predicate in a given context, etc. These days, I am thinking more in terms of "shapes" (see SHACL and ShEx as examples of shapes languages built on RDF), which are more like YAML or JSON where you are free to give a note an arbitrary structure, but if the structure conforms to a schema, there are additional benefits w.r.t. queries and inference over your knowledge graph. If Dragon (https://www.meetup.com/Category-Theory/events/277673652/) becomes open source, then Dragon will become the data and schema language for the next version of SmSn. I am less interested in Markdown now than I was when I wrote the above. YAML offers more opportunities for explicit structure, and is even friendlier to line diffs.