Terminology Notes: Archive, Register, Index, Repository

Four words come up repeatedly in conversations about small collections of technical reference material: archive, register, index, repository. They are sometimes used interchangeably and sometimes with very specific distinctions. This note records the working distinctions I use in my own practice.

Archive

An archive, in the sense used here, is a collection of records preserved with attention to their continued accessibility but not necessarily organised for browsing. Entries enter the archive individually, are annotated according to a small fixed schema, and remain in place. The principal use is retrieval-on-demand rather than browsing; the principal discipline is custodial care of stable references.

The word carries a connotation of long duration that I find apt. An archive is a collection that is expected to outlast the circumstances under which it was created, and the discipline of maintaining one operates on that expectation.

Register

A register is a flat list of entries, typically maintained for the purpose of having a complete enumeration. Registers are common in administrative contexts (a register of patents, a register of births) where the principal use is to confirm or deny membership rather than to retrieve substance. The register's emphasis is on exhaustiveness within its scope.

For technical material this emphasis is rarely the right one. A personal archive is by definition partial; the items not in it are not absent because they have been excluded but because they were not candidates in the first place. Calling such a collection a register would mislead the reader into expecting exhaustive coverage.

Index

An index is a derived structure: it points at material that exists elsewhere. The index does not itself constitute the material. A library's index is the apparatus that allows the library's holdings to be searched; it is not, itself, the library.

This distinction matters because a personal archive, in the sense used here, is a collection of entries rather than a collection of pointers to entries. The substance of the archive is in the entries themselves. To describe the collection as an "index" would suggest that the substance is held elsewhere, which in this case it is not.

Repository

A repository, in technical usage, is a storage location for material that is being actively edited or operated on. The term is common in software contexts (source code repositories) where the expectation is of frequent modification and a working history of revisions. A repository is, in this usage, somewhere material lives during its working life, not somewhere it goes when it is settled.

For settled reference material the connotations of "repository" are slightly off — they suggest live edits where the present case calls for archived stability. The repository terminology does fit some of the supporting infrastructure of the archive (the place where draft entries live before they are accepted into the archive proper), but not the archive itself.

The choice

The conclusion of all this is that "archive" is the closest fit for the present collection. The collection consists of entries preserved for stable reference, organised with attention to durable identifiers (opaque hex strings), annotated with a small fixed schema (title, dates, tags), and not expected to be browsed. None of the alternative terms captures this combination as cleanly.