Git tracks file identity of annexed content (data), so anyone cloning the repository could potentially access metadata of data. Is this ok?
There is no elegant solution (so far) for tracking data in datalad but ignoring via git, since datalada utilizes git for its infrastructure. It's like trying to cap an engine's maximum MPH while still attempting to ensure that the car could go past that cap.
Not only that, but datalad permissions architecture further complicates the data provenance workflow
All these problems highlight the need for further planning data-monitoring, and studying of Datalad + Git infrastructure. I propose a week.
Challenge
All these problems highlight the need for further planning data-monitoring, and studying of Datalad + Git infrastructure. I propose a week.