Feature Request: Incremental (PATCH/Append) API method for Experimental Data Files

[adndvlp]

Hi OSF Team,

I’m currently working with jsPsych and DataPipe in behavioral experiments. The OSF API only allows uploading full files (CSV, JSON), which makes it difficult to save trial-by-trial data efficiently in long experiments.

Purpose:
The main goal is to save results incrementally in case a participant abandons an experiment, ensuring that previously submitted data is not lost. While DataPipe could leverage this new API method to handle incremental updates, other services using OSF could also benefit from this capability.

Proposed Solution:

• Provide an incremental (PATCH/append) API method that updates an existing file by adding new information for each "stimulus-response" trial, rather than uploading the entire file again. Use file IDs to make locating and updating data fast and straightforward.

This would allow sending only new trial data without re-uploading the full dataset, improving efficiency and reducing server load.

I understand this may require significant architectural changes, but I hope this feature can be considered, as it would greatly benefit researchers managing large or long-running datasets.

Thank you very much for considering this suggestion!

Best regards,
Andrés