Diaryx is inspired by the Joseph Smith Papers Project, which is described on its website as:

an effort to gather together all extant Joseph Smith documents and to publish complete and accurate transcripts of those documents with both textual and contextual annotation1

Diaryx enables anyone to have their own “Papers Project.” It is designed to make archival, annotation, and publication accessible to anyone.

What Diaryx borrows

Diaryx did not invent the shape of its data model. The archival and records-management professions formalized most of it decades ago, and where a formalization exists Diaryx follows it rather than a fresh guess. What follows is the lineage, so that a reader can check the work against its sources.

One corpus, many arrangements. A METS document holds a single flat fileSec and any number of typed structural maps over it — a physical map walking boxes and folders, a logical map walking chapters or correspondence — each pointing at the same files.2 No map is the real one. Diaryx's lenses are that idea: the containment spine is one view over the corpus rather than the corpus itself, and a document filed under a year is the same document that appears under a person and a place.

Classification is not aggregation. ISO 15489 defines classification as identifying a record by the context that produced it. MoReq2010 separates that from aggregation, "the activity of assembling related records together," which "may be based on any organisational requirement or criteria, not business context alone."3 Diaryx keeps the same seam: where a file sits and how a view groups it are two decisions, and a view's grouping never dictates a folder. MoReq2010 permits welding the two together and warns against it — organizations that do end up hybridizing their scheme, mixing subject-based categories into a structural one, or splitting apart aggregations that belonged together. A view should never become a reason to move a file.

Declare the scheme, inherit it, allow the exception. MoReq2010's hierarchical classification module applies a class at an anchor, inherits it as the default for every descendant, and lets a class applied directly to a child break that chain — which is how a collection holds records that don't all share its classification.3 A Diaryx view names its anchor, covers the subtree beneath it, and yields to a document that says otherwise.

Records are a graph, not a tree. Records in Contexts replaces the single fonds hierarchy — an artifact of what a paper finding aid could physically be — with entities and relations: records, record sets, agents, activities, places, dates, and membership that is not confined to one parent.4 Diaryx's typed relations and its refusal to make the folder tree the only structure are the same correction.

Arrangement is documented, not inferred. A finding aid states its arrangement scheme explicitly, because a scheme reconstructed by guessing at patterns is exactly what the profession learned to distrust.5 So Diaryx resolves placement from a declaration in the vault, never from a filename or a title that happens to match.

Fixity is an archival obligation. PREMIS treats checksums and the events that verify them as preservation metadata a repository owes its holdings, not as an optimization. Diaryx checksums document bodies by default for the same reason.

Where Diaryx departs from all of these is in who is doing the work and when. Every standard above describes a closed corpus, held by an institution, arranged after the fact by someone other than its creator. None of them are written for a person filing tomorrow's entry into their own papers. That is the gap Diaryx exists in: archival-grade description applied at the moment of writing, at the scale of one person, with the structure carried in the documents themselves rather than in a separate instrument that describes them.

  1. https://www.josephsmithpapers.org/articles/faq1#question-1 ↩

  2. METS, the Metadata Encoding and Transmission Standard — https://www.loc.gov/standards/mets/ ↩

  3. MoReq2010® Volume 1, Core Services & Plug-in Modules, §1.4.5 (classification and aggregation) and §201 (hierarchical classification) — https://moreq.info/files/moreq2010_vol1_v1_1_en.pdf ↩ ↩2

  4. ICA Records in Contexts — https://www.ica.org/ica-network/expert-groups/egad/records-in-contexts-ric/ ↩

  5. Describing Archives: A Content Standard (DACS), Part I — https://saa-ts-dacs.github.io/dacs/ ↩