queue_isBoundedAndDropsRatherThanGrows flakes on Java 8: one failed loopback send fails the drops-are-silent assertion #22

main is red: the push run for cde8148 (#21), https://github.com/Stephenson-Software/trace-client-java/actions/runs/39232050979, failed on the build (8) leg. The 17 and 21 legs were cancelled.

TraceClientTest.queue_isBoundedAndDropsRatherThanGrows:518 a full queue drops without logging, but logged: 1 ==> expected: <true> but was: <false>

This is a flake, not a regression. #21 changed only Javadoc, and the same tree passed both of its pull_request runs minutes earlier. The log.records.isEmpty() assertion was added in #20.

Cause, as far as can be told: the server in that test only answers 201, so the one line the send path can log is could not deliver …, which send() writes when an IOException hits one of the ~257 short-lived loopback connections. The assertion assumes none of those connections ever fails, which loopback does not guarantee. The failure message prints only the count, so the line itself was lost.

Expected fix: the test keeps its regression value (a client that logged each drop would write hundreds of lines) but tolerates the odd failed send, and its failure message prints the logged messages.

This issue was drafted during a Gardener session (https://github.com/Stephenson-Software/gardener).

drafted by Claude on behalf of Daniel Stephenson