The recent change of the Resource to represent family history information in the continuous integration build ( previously called FamilyHistory and now called FamilyMemberHistory) from an object to a generic List type is a major step backwards. Assertions that are made regarding family history are more than just those made about a given relative ( previously called relation). A FamilyHistory FHIR resource needs to represent more items than making assertions regarding individual family members ( now called FamilyMemberHistory). The attributes that are lost include being able to make use of extensions to handle negation, the timestamp on when the family history was taken, assertions and interpretation made at the level when interogating the entire family, attaching a pdf or svg to represent the pedigree, assertions regarding the ancestry, results of clinical decision support algorithms using extension, free-text notes and/or comments made about the entire family, fall-back HTML. While some of these things can be handled with pushing items onto the new list structure, such as an attachment resource for the pedigree, other resources don't make sense. For instance, using an Obversation Resource to handle the negation seems semantically wrong. Obversations should be about items the clinician... well observes. Family history are assertions made from the patient and are hearsay in nature and should not be held to the same weight as 'true' observations made by a clinician. Finally , I am also not sure how this change was made without discussion with the Clinical Genomics working group, which I am on. Unfortunately, I won't be able to argue this point in person in Paris.