Inconsistency in naming of downloaded TSV files
If I go to a workspace and download the metadata for the participant, sample and pair entities, FireCloud generates participant.txt, sample.txt and pair.txt files. These are the TSV-formatted files that can be used for populating a workspace's data model. If I download the metadata for a set entity (e.g., a pair set), FireCloud generates a zip file (e.g., pair_set.zip) containing two TSV files: a set_entity.tsv and a set_membership.tsv. For example, in the case of a pair_set, the zip file contains two files: pair_set_entity.tsv and pair_set_membership.tsv. The naming convention for the set entities is different than the naming convention for the non-set entities: the former have the tsv file suffix, while the latter have the txt file suffix. The naming conventions should be consistent.