Skip to content

Repository cleanup: dead packaging, uncollected tests, stale docs - #176

Merged
anderslindho merged 9 commits into
masterfrom
chore/cleanup
Aug 7, 2026
Merged

Repository cleanup: dead packaging, uncollected tests, stale docs#176
anderslindho merged 9 commits into
masterfrom
chore/cleanup

Conversation

@anderslindho

Copy link
Copy Markdown
Contributor

Housekeeping accumulated since reccaster moved out, one commit per concern.

Two worth a look before the rest. The debian packaging is removed; it cannot
build, but it is the one change with reach outside the repository, so say if any
site still builds from it. The README changes describe current state only and
leave the framing quoted in #174 unchanged, so that question stays open.

DS_Store was missing its leading dot and `.*.swp` did not match a bare
`.swp` or `.swo`, so both kept showing up as untracked for anyone without
a global ignore file covering them.
The clang-format hook only ever applied to the C sources under `client/`,
which have moved to ChannelFinder/reccaster. The remaining comments came
from the upstream hook templates, and the file was missing the trailing
newline `end-of-file-fixer` expects.
UTF-8 is the default source encoding in Python 3; these are left over
from when the tree also supported Python 2.
The entry still read `docker-compose.yml`, so it matched nothing and the
compose file was copied into the build context.
It cannot build. `debian/rules` drives an EPICS build under `client/`,
which is now a stub, and the packaging targets Python 2.7 throughout.
Nothing in the repository or in CI references it, and the sqlite3
initialisation that `debian/README` documented is also in
`server/README.md`.
`server.yml` addresses `tests/unit`, `tests/integration` and
`tests/container` individually, so the two files at the top of `tests/`
ran under a local `pytest tests` but never in CI. `test_recast.py` is
renamed on the way because the name is taken below `unit/`.
It documents the same options as `demo.conf` and had drifted:
`commitSizeLimit` was missing, and `pushAlwaysRetry` lacked the note that
enabling it blocks commits for every IOC until ChannelFinder recovers.
The integration IOC services wait on `recc` being healthy, and what they
need is RecCeiver's TCP listener. The check looked for an established
connection to ChannelFinder instead, which reflects `requests` connection
pooling rather than whether RecCeiver is up. `timeout` was also longer
than `interval`, so probes could overlap.
@anderslindho anderslindho self-assigned this Aug 5, 2026
@sonarqubecloud

sonarqubecloud Bot commented Aug 5, 2026

Copy link
Copy Markdown

@tynanford tynanford left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good to me. Sounds like the debian packaging hasn't been maintained for awhile. Looks like FRIB might have added/maintained that before. Tagging @acar7901 in case

@anderslindho
anderslindho merged commit 549b1db into master Aug 7, 2026
15 of 16 checks passed
@anderslindho
anderslindho deleted the chore/cleanup branch August 7, 2026 11:52
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants