diff options
| author | Pjotr Prins <pjotr.public01@thebird.nl> | 2026-10-05 07:47:10 +0000 |
|---|---|---|
| committer | Pjotr <wrk@bosa.lan> | 2026-10-05 07:47:10 +0000 |
| commit | 8140dc2a8d7e24f4e530bd471819d18c66b732ee (patch) | |
| tree | a82000741e7f0ff8f2fa0af1df6770bdaadb16fe /test/python/server_test_case.py | |
| parent | 551bb2f8a6aaa420b6bdd4729e4fb2bd2f739b5d (diff) | |
| download | genenetwork4-8140dc2a8d7e24f4e530bd471819d18c66b732ee.tar.gz | |
doc: Fix staleness in README.md and RELEASE_NOTES.md
Three claims left over from before "every endpoint requires a token"
(previous commit) were actually wrong, not just stale phrasing:
- RELEASE_NOTES.md's intro still said "not all endpoints will require a
token" -- the opposite of what the rest of the file by then described.
- README's /traits/{dataset} entry and RELEASE_NOTES.md's /traits/...
entry both still said /groups/{group}/traits "isn't gated" -- true
for resource-level authorization, false for the token requirement
added in the same commit that should have updated this.
Also fixed real staleness independent of this turn: README's Tests
section enumerated specific Rust/Python test file pairs needing a live
database, a list that had silently fallen behind several rounds of new
endpoints (missing species_datasets, group_traits, dataset_probesets,
traits, get_test_token). Replaced the enumeration with a description of
which tests need what and why, so it can't go stale the same way again.
Also documented AUTH_SERVER_URL/GENENETWORK4_DEV_MODE/GENENETWORK4_BIND
for the "run the test suite with a self-provisioned token" path the
previous commit added, which the cargo test / python -m unittest
invocations here hadn't mentioned yet.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XDDpXvZBQat9uSgDeEMJ8d
Diffstat (limited to 'test/python/server_test_case.py')
0 files changed, 0 insertions, 0 deletions
