Description
An emerging concern around media files is whether or not users will understand how to fetch the correct image info to use in attribution. If an image URL is fetched from the article directly, it reflects an upload URL that can not be used directly in attribution requests. Instead, users are expected to transform the URL to an actual commons request, or use commons directly.
Because we don't actually know how images are being fetched by our users, we do not know how big of a problem this is. In addition to consulting with partners and users directly, we should implement monitoring to better understand usage patterns in practice.
Conditions of acceptance
- Add a chart to grafana that differentiates media file failure types for 404 errors:
- Article lookup URL (eg: starts with 'upload') is not supported
- Poorly formatted page title (eg: our wonky URL formatting isn't correct)
- Requested page actually doesn't exist