2022-02-21

Standup

Last

• Anu: did not work
• Rising: did not work
• Leo: with Rufus this morning
• Rufus:
  • finished PRODUCT page with initial product roadmap pages
  • Plan of work for v0.1.1 pages-v0.1.0-readme-with-csv-table-preview
  • Data-Literate example now here https://github.com/datopian/portal.js/tree/main/examples/data-literate

Next

• Rising: unavailable
• Anu: unvailable
• Leo: ~1/2
• Rufus: split out tasks to team, keep progressing on data literate demo

Blockers / to be discussed

• Resourcing

Notes

Rufus

Goal: Planning Stream document processing complete and PRODUCT done and PLAN started

• review 2022-02 Consolidate the Analysis Project to see where we are at
  • IMO priority is getting features in and a clustering and options (and/or choice) into pages
    • Qu: (ask Leo) - do i move the feature coggle into a spreadsheet *Leo: Keep in coggle for now and just move the stuff we need
      • Why? so that we can add filters, add more comments, track status etc
      • Why not? Not worth the hassle
    • Qu: should we use job stories or features? Leo: feature with clear goal may be better than job stories for now
      • Features is more requirements-y vs job stories and job stories are better …
      • When I have a CSV of data I want to view it/display it so that I or others can see what the data looks like
        • What happens if the data is really big …
        • Text-wrapping …
        • Column sizing and resizing

Asides

• branding discussion. My sense is we go all in on DataHub and even brand portaljs in that direction e.g. DataHub Toolkit or similar (??)

Outflow re pages MVP (before i get into the trees)

What do i want as a user?

• Scratchpad (to publication) experience: I want to move as fast as possible. In particular, i want a preview as I play around (locally) and potentially be able to share that very quickly
  • Is it key i can preview locally before deploying? 💬 for me Rufus probably yes, or rather i want to do this
• Dataset publishing experience: I've got the metadata etc and i want

Things we could build

• Self-service using the app: i.e. npm install, run locally and then setup github pages
  • ➖ suspect there is an issue in UX in terms of installation complexity (and issues) and speed. npm often goes wrong (e.g. going wrong for Leo) and users will just give up
  • 💡? could build a plugin into obsidian and going with that rather than running npm https://github.com/obsidianmd/obsidian-sample-plugin
• I have stuff locally but i can connect to an API that renders and gives me a preview. Like https://github.com/joeyespo/grip but for data stuff.
  • Walkthrough
    • run datahub preview (which keeps running in watch mode)
    • it sends my README (and data) to pages.datahub.io/api/preview
    • returns a url like pages.datahub.io/previews/my-random-url which contains preview
  • Issues: data stuff may be a problem esp if data is large … (?)
• I start playing online straight away, e.g. it is like opening a github issue or hackmd and i just keep uploading stuff straight into an editor page and it gives me a live preview (and stores my data)

Still feel self-service is the way to start since we need it for other items … (it runs through the core codebase)

However, wonder whether sufficiently good UX for people outside of the core team