Wait for the wrong diff task
After posting /diffs, the response includes the Celery task id under uid, yet create_diff ignores it and instead fetches /tasks/active once and waits on whatever task happens to be at index 0. If the new job has not started yet or other jobs are already active, TASK_ID may be empty or belong to another dataset, leading to a premature "No active tasks found" error or polling the wrong task while the intended diff continues running. The CLI should poll /tasks/{uid} for the id returned by the create call rather than relying on the first active task.
Useful? React with 👍 / 👎.
Originally posted by @chatgpt-codex-connector[bot] in #131 (comment)
After posting
/diffs, the response includes the Celery task id underuid, yetcreate_diffignores it and instead fetches/tasks/activeonce and waits on whatever task happens to be at index 0. If the new job has not started yet or other jobs are already active,TASK_IDmay be empty or belong to another dataset, leading to a premature "No active tasks found" error or polling the wrong task while the intended diff continues running. The CLI should poll/tasks/{uid}for the id returned by the create call rather than relying on the first active task.Useful? React with 👍 / 👎.
Originally posted by @chatgpt-codex-connector[bot] in #131 (comment)