Rufus

• Create ../design-showcase
• Create ../design
  • Outline
  • Move domain model stuff there
• Sketch out two paths: Data Project vs Data Scratch(pad)
• Markdown Pipeline notes ../markdown-pipeline ✅2023-03-05 good start

Inbox

• Create journal stuff in datahub next or not?
  • What does that mean? ✅2023-03-05 it means having a journal folder and creating journal notes
  • Why / why not?
    • 👍 IME using journal as default location for notes is really useful b/c
      • reduces cognitive load of deciding
      • allows one to start braindumping immediately
      • provides a default capture location for tasks
    • 👎 messy and pollutes this repo with random thoughts / notes.
      • Against this: encourages us to work in the open, yay!
      • Plus can always just keep notes in repo and not publish them.
  • Choice ✅2023-03-05 let's go for it
  • Next steps ✅2023-03-05
    • Set up journal folder
    • Add first daily note (today!)
• The great big issue tree in the sky? A consolidated question tree? ✅2023-03-07 see ../brainstorm-2023-03-06.canvas

Scratch

Pipeline workflow

Use cases

• rendering a table of contents for the given page on RHS (do you do it at level 1 or level 2). Also note how this is rendering content on the page outside of the main text body.
• showing backlinks and forward links
• image assets

Blocks are the way to go

#todo/process

CMS have gone all block-y i.e. they use pages composed of blocks e.g. wordpress editor, notion.so etc. In a way blocks are just the good ol' dom nodes but with metadata and associated editor UI.

You also see this in headless editor work where they tend to talk about components.

We want to adopt this i suspect.

Concrete example: imagine a README.md with a dataset embedded in it somehow e.g.

# My cool dataset

<TabularData src=""

Not quite getting the point across.

Notion model

See ../notion-domain-model

Random

• https://github.com/errata-ai/vale - linting for prose #links