File naming for personal archives

Start with identifiers, not clever names

A useful filename is short enough to handle, unique within its context and understandable without opening the file. Use a stable identifier, an optional date, a brief description and a sequence or role. Document the convention before renaming a whole collection.

Keep names consistent. Use leading zeros for a fixed sequence, such as 001 rather than mixing 1, 02 and 003. Use a date only at the precision supported by the evidence. A year is more honest than inventing a month and day.

Record the old name

Renaming can break links in projects and catalogues. Keep a mapping of old path to new path. For collections that depend on software, preserve a complete original export before changing directory structure.

Keep the actual file extension. Changing .jpg to .tif does not convert an image. A filename labels a file; an application performs a conversion and may change the content.

Trial the convention

Try ten representative files: an unknown date, a long surname, a multipage document, an edited image and an item with a companion file. Check for collisions and sorting surprises. Only apply the rule more widely after the trial makes sense.

Use descriptive metadata for detail that would make a filename unwieldy. Multiple people, alternative place names and disputed identifications belong in an inventory or metadata record connected by the identifier.

Write instructions for the next custodian

Explain code meanings, sequence rules and the date convention in a README file. Specify where original filenames are recorded and how derivatives are named.

The practical test is retrieval: give someone an inventory identifier and have them find the matching file. A highly structured filename has little value if no one can interpret its parts.

Working checklist

Use this as a session checklist. Ticks are saved only in this browser; download a worksheet to keep a portable record.