batch_evaluation fails on self-hosted v4 events_only deployments (uses unavailable v3 read endpoints)

Environment

• Langfuse server: self-hosted v4 (docker.io/langfuse/langfuse:4), default write mode events_only
• Python SDK: 4.14.4 (also verified against 4.15.1 source)

What happened

Running batch evaluation against existing traces fails. The runner fetches items via endpoints that are unavailable on v4 events_only deployments:

• scope="traces" → client.api.trace.list → GET /api/public/traces
• scope="observations" → legacy observations_v1 → GET /api/public/observations

Both return:

"This endpoint is not available on deployments running in Langfuse v4 events_only mode."

Root cause

BatchEvaluationRunner._fetch_batch_with_retry (langfuse/batch_evaluation.py, line 1139 in v4.14.4) still calls the v3 read APIs. Per the v3→v4 migration guide, reads should go through GET /api/public/v2/observations (client-side grouping by traceId, root observation for trace-level io).

Expected

batch_evaluation works on events_only deployments, e.g. by fetching via the v2 observations API.

Workaround

We currently run our own thin loop: fetch v2 observations grouped by traceId → evaluate → create_score (score-create ingestion works fine on v4).

Happy to contribute a PR if the v2-based approach sounds right — glad to align on the design first.