I would like to come back to a discussion during the Nov. 8th, 2024 virtual community meeting:
It was mentioned by @juanf that datasets would dissappear from the UI and files would become directly associated with objects. To us this would be an extremeley unpleasant situation. We rely on the semantics and functionality of datasets:
- We distinguish dataset types for raw data, processed data, machine settings documentations etc. and rely on the dataset type to identify the correct data in external processings.
- Datasets may contain large number of files which clutter the view if not grouped by datasets.
- Dataset metadata is sometimes relavant for understanding the data in the dataset correctly.
Please consider these arguments in your developements.
Thanks and best regards.