You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Commit 698f341
Browse filesBrowse the repository at this point in the historyBrowse files
Storage release after a pass fails the test again; mark the invented fake answer (#189)
Review round 43: in the two tests whose cleanup now warns instead of throwing,
a release that fails after the test's own checks passed fails the test, as on
main, while a failure above still stays the one reported; the second holder
fake case says its answer is invented.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
// Answers as Docker 27.4 does for one swept network with two candidate holders, one gone before the inspect: the
149
149
// inspect prints the other and exits 1. Every other call gets no output and exit 0, as for `ls` with no match; the
150
-
// removals' stdout (the removed IDs or names in 27.4) is not read. The network line holds only the fields the sweep
150
+
// removals' stdout (the removed IDs or names in 27.4) is not read. The second case's answer is invented: it stands in
151
+
// for an unexpected inspect failure that still printed some holders, which 27.4 is not known to give. The network line holds only the fields the sweep
151
152
// reads (27.4 also prints CreatedAt, Driver, IPv6, Internal, Labels and Scope), and the inspect an empty endpoint for
152
153
// the network, where 27.4 gives its full settings: the sweep reads only the network names.
0 commit comments