Closed redmitry closed 1 year ago
For the "participant" datasets there are situation when the Participant (tool) has no inputs (simulations, synthetic data generatiors, etc). This way the dataset must be able to have the dependence on the tool only. Current JsonSchema disallows this: https://github.com/inab/benchmarking-data-model/blob/0c92a927d3367dffea1a238aadd2d71c85848b40/json-schemas/1.0.x/dataset.json#L225
we need either make rel_dataset_ids optional or IMHO better remove minItems: 1 restriction (which is more explicit).
@redmitry could you check whether change at 2786fbd6febaac5b050f1b1b5a4be3dbb352f195 could cope with the use case, please?
As no feedback was received in these months, I consider the use case is properly covered
For the "participant" datasets there are situation when the Participant (tool) has no inputs (simulations, synthetic data generatiors, etc). This way the dataset must be able to have the dependence on the tool only. Current JsonSchema disallows this: https://github.com/inab/benchmarking-data-model/blob/0c92a927d3367dffea1a238aadd2d71c85848b40/json-schemas/1.0.x/dataset.json#L225
we need either make rel_dataset_ids optional or IMHO better remove minItems: 1 restriction (which is more explicit).