This directory holds repository-only evidence for Alan Dipert's open-source
work. It is intentionally outside md/ and is never copied to the generated
website.
repositories.jsonl and gists.jsonl are snapshots of the GitHub inventory.
gist-index.tsv makes every gist reviewable, while gist-footprint.tsv records
only downstream uses supported by direct attribution or use reports.
repository-index.tsv is the readable 199-row catalog. Its relationship
column says owned or affiliated; affiliation is a preservation lead, not
an authorship claim.
Complete Git bundles live below git/; known-upstreams.txt lists canonical
repositories that must be preserved in addition to Alan's repositories and
forks. contributions.tsv is the evidence ledger behind the curated public
list. It distinguishes authorship, preservation, independent contributions,
and paid open-source work. Commit counts are capture-time snapshots linked by
GitHub to Alan's identity, not a complete authorship census.
The inventory and bundle commands are explicit network operations:
python3 tools/archive_personal.py fetch-github-inventory
python3 tools/archive_personal.py bundle-github
python3 tools/archive_personal.py fetch-gist-footprint
python3 tools/archive_personal.py fetch-sourceforge-jd4x
python3 tools/archive_personal.py checksums
make archive-audit
The bundle command is resumable. It verifies and skips an existing bundle
unless --update is supplied. bundle-failures.json records remote
repositories that could not be fetched, and a nonempty failure list makes the
command fail.
The full corpus is large. Git bundles, WARCs, release archives, and media are tracked with Git LFS. Initialize Git LFS before materializing it.