Unfortunately it’s not very user-friendly to develop datasets & narratives in tandem - this is a good reminder for us to update the docs here (or better still, improve auspice).
The domain for datasets defined in the markdown (either localhost:4000 or nextstrain.org) isn’t actually important, rather it’s the path that matters. The Auspice client, when parsing the narrative, uses this path to make dataset requests to the domain it (auspice) is served from - the domain specified in the markdown file is ignored.
As an example, using your narrative markdown file which defines the dataset
- If you are using auspice locally, e.g. via
auspice view on localhost:4000, then the dataset http://localhost:4000/ncov/bavaria must exist.
- If you are using auspice through nextstrain.org, then the dataset
https://nextstrain.org/ncov/bavaria must exist. This doesn’t exist, which is why you can’t view your narrative through the community functionality.
More information available in this GitHub issue