The scientific papers (Articles) are stored in a knowledge graph. The tool used is Weaviate.
To start this application first build it:
$ docker-compose build --no-cache
# run all the services
$ docker-compose up
Note: Once the download (+/- 1.6 GB) of the instance is done, your local Weaviate instance will install. This might take a while. Once it's done you can run it on http://localhost:8080/v1/meta
.
Note: also anytime the config, data or client changes re-run previous command in the corresponding directory
Now you can open the client in the browser window by opening http://localhost:3000/
in your browser.
- User Interface:
./client
- Server (for Mailchimp subscription):
./server
- Articles (Weaviate):
./weaviate
Deployment is managed via Github. Once a push has been done an automated task will deploy this project. A typical deployment will take about +/- 15 mins.
Bug: Occasionally the build will fail. One example is that the disk space is full. (json.log: no space left on device). Since this is managed elsewhere this repo has no control over this and will not be notified...