Open kevinjcliao opened 3 years ago
I'll start with automated deployment, then see if it's possible to run without server.
Now that people are trying to figure out what features will be necessary in the foreseeable future, I think it's better to think conservatively in terms of architecture -- maybe we will still need a server. Although for now all the data I can imagine are static-ish and can be refreshed periodically.
Ok, at least I got the scraper working on GitHub Actions now. It produces a JSON and pushes to gh-pages
branch (which can eventually be merged into React.js front-end code)
The next challenge is to extract static code and deploy to GitHub Pages:
GET /hospital
endpoint with a JSON that is a combination of hospitals.csv
and availability.json
. (hard)
hospitals.csv
into hospitals.json
then let the frontend to merge & display two dataset, then it'll be easier to manage data. hospitals.json
can be generated as a part of static site generation. Its change rate is much less than availability.json
.https://github.com/yorkxin/vaccinate/blob/gh-pages/data/availability.json
On-prem management is exhausting and we don't need it. @yorkxin is exploring alternatives.