Re-examining the remaining bits of the db structure that I haven't worked on yet, it seems to me that the relationship between documents, files, and document-to-inscription relationships is very puzzling. Either I've failed to understand it properly, or we've planned it badly. I suspect that what we really need are a separate documents/doc-to-insc component, and a file/file-to-insc component; or that the file, commentary and transcription fields should actually be part of the main inscription record. Waiting for info from GR to make decisions on this.