2022-03-15
2022-03-15
Plan
- data-literate reverted
- next v12
- TableGrid back (and in example)
- Remove data-literate-template for now
Demo repo up
- Demonstration repo up
- Create repo with MIT license
- Add content from
your-own-data-example
- Script to do build in next.datahub.io repo
- Build in github pages
- Build is failing: https://github.com/datopian/datahub-pages-demo
- Aside: actions@v3 is the latest (we are currently using actions@v2 in deploy script)
- Build is failing: https://github.com/datopian/datahub-pages-demo
- (FUTURE) Build in vercel/cloudflare etc
- just deploy to a branch and setup vercel etc to build from that branch and NOT do next build && next export in the workflow script
Tidying data-literate
- What about the "irrelevant" files in data-literate?
- Data files in
public/_files? ✅ keep for now - though want to replace the excel file with something smaller in future- Add instructions to the tutorial about removing? ✅2022-03-15 WONTFIX for now as not a big deal
- your-own-data-example => that move out to a separate repo (it's the demo we talked about)
- Data files in
Next
Update the tutorial …
- Fix link to issue tracker because current one is private
- Should we make a public project for this?
- What would be in it? e.g. this could include the build script? Or it could include everything (except docs)?
- What are the pros and cons?
- I like to default to open
- Add screenshots (?)
- Deployment: are we assuming people have data already in git(hub) or just on disk?
- If not we should add that as a first step (though with vercel you can deploy without pushing to github)
- But all the others require git(hub) so maybe we should just assume that
- This makes me think about the general context of the tutorial … => maybe need to flag that a bit more (do you already have data in git(hub) or just some data lying around on disk)
- If not we should add that as a first step (though with vercel you can deploy without pushing to github)
- Can we have some real world data that's useful / interesting?
Notes
#done/moved to tutorials
Think about what goes in a second tutorial
- What features do we people want that we can showcase
- How can we eat our own dogfood?
Rufus brainstorm
#done/moved to INBOX and 2022-03-16
- datopian.com - https://github.com/datopian/www.datopian.com
- Publish another datasets or two
- Tell a data story
- Make a default dataset page that looks awesome
Other things to do:
- try out tina.io (as alternative to contentlayer)
- write up notes in docs
What's our focus and priority?
#done/moved to INBOX
Approach: Move fast, release often, share publicly
- Putting awesome data on datahub again … so that we practice running a hub, because the data is awesome
- Aside: Let's go full github ;-) e.g. using commit history and all that stuff
Misc: take down sign up at datahub.io and instead have email signup form for the "New Coming Soon Next Version of DataHub" (or something like that 😉)
Asides
- portaljs.org => portaljs.datahub.io or portaljs.datopian.com?
- script to process "your repo into nextjs" can do some useful things e.g.
- process all markdown to one place all non-markdown to another place … so that you can keep stuff together e.g. data next to content (??)
Examples:
- https://github.com/datasets/rio2016 - how could this become a nice web page
- Having structured metadata auto displayed rather than manually entered
- Having preview of the data etc
