packageR

packageR started as a fork of the excellent File Browser project. Most core File Browser concepts still apply: users, permissions, shares, a familiar file tree, and browser-based inspection of text and common media.

packageR is meant for data that lives in object storage and is made visible to packageR as a regular file tree. That filesystem view can come from different deployment choices, where an object-storage bucket is mounted into the runtime, often through FUSE or the platform's storage integration.

On top of that filesystem view, packageR adds bucket-aware features: presigned URLs for direct access, streaming previews for supported formats, and catalogs for domain workflows. STAC is the main geospatial catalog workflow.

Large uploads and downloads should usually go directly to the bucket with S3-compatible tools. packageR is mainly for browsing, inspection, curation, sharing, preview, and catalog access.

Another conceptual difference is operational. packageR treats the File Browser database as ephemeral cache: only files in the mounted filesystem are persistent, while runtime state such as configured shares comes from startup configuration. This lets operators treat packageR as stateless, with no application database to back up; the persistent system is the mounted object store plus credentials for presigned URLs. A curated init.sh bootstrap provides recommended defaults and helper flags, while still letting deployments tweak settings through environment variables.

Use the feature summary for the packageR-specific behavior layered on top of File Browser.

The quickstart starts a local sample. The walkthroughs also cover public sharing, presigned URLs, access boundaries, and STAC catalog access.