NEWS

---

researchR 0.1.2

• Keep source API keys and bearer tokens out of displayed request errors, including validation responses from older servers that echo request input and JSON-escaped or URL-encoded credentials in diagnostics.
• Serialize named concept lists, recipient/delivery vectors and scrape-target lists as JSON arrays, preserving their order and values.
• Reject malformed API metadata, missing row collections, invalid structured metadata and non-finite date bounds instead of reporting a successful connection, empty metadata or a misleading filter result.
• Prevent package checks and ordinary example runs from executing the examples that create, delete or reconfigure questions. Clarify feed limits and errors.

researchR 0.1.1

• Preserve the time, fractional seconds and UTC offset of server timestamps. Empty question, health and heartbeat responses now retain their full schema and column types.
• Validate connection settings, question definitions, filters and pagination before requests; encode legacy question IDs safely in paths and feed URLs.
• Detect inconsistent pagination instead of returning silently incomplete or duplicated results. Distinguish temporary HTTP 503 responses from a disabled API, and render structured server validation errors without interpreting their braces as formatting expressions.
• Return record provenance as a list column while retaining compatibility with older servers, and clarify first use, feed limits and collection boundaries.

researchR 0.1.0

• Standardize the named v5 package logo across README, documentation pages and available application headers, with matching light and dark artwork.

• Added an executable offline getting-started vignette covering connection configuration, concepts and feed URLs, with explicit boundaries for server requests and email delivery.

• Added a grouped reference index using the shared CTTIR documentation theme.