Manifests: Pinning a Dataset Too Large to Copy by Listing Objects and Their Digests
The outcome explicitly concerns data far larger than one machine holds, and a result cannot be re-derived if its input set is undefined. This is the reproducibility answer to scale, and it is what binds a whole run's input set to identities that outlive paths - something object versioning cannot do, since versioning only pins one store's own generations of one key. It is also the node that makes the immutable-object model matter to the researcher rather than only to the bill.
This Concept is waiting for its first lesson!
The outcome explicitly concerns data far larger than one machine holds, and a result cannot be re-derived if its input set is undefined. This is the reproducibility answer to scale, and it is what binds a whole run's input set to identities that outlive paths - something object versioning cannot do, since versioning only pins one store's own generations of one key. It is also the node that makes the immutable-object model matter to the researcher rather than only to the bill.
Are you a teacher? Sign in to start contributing.
Sign In