Closed drewnoakes closed 4 years ago
@somdoron I cannot see why this test is failing. Is it known to be flakey? I see quite a few PRs merged with failing tests.
NetMQ.Tests.XPubSubTests.MultiplePublishers()
X NetMQ.Tests.XPubSubTests.MultiplePublishers [509ms]
Error Message:
Assert.Equal() Failure
↓ (pos 23)
Expected: ···sage from subscriber
Actual: ···sage from subscriber 2
↑ (pos 23)
Stack Trace:
at NetMQ.Tests.XPubSubTests.MultiplePublishers() in /home/runner/work/netmq/netmq/src/NetMQ.Tests/XPubSubTests.cs:line 279
yes, it is flakey...
LGTM, I think the compiler can figure out some of the assumes itself. You can merge :)
Please especially review the few public API changes towards the end.
Note that
Assumes.NotNull
is[Conditional("DEBUG")]
so has no runtime cost in release builds. I try to avoid using!
where possible, and new code should be designed such thatAssumes.NotNull
isn't needed either. It's not reasonable to change the design on existing code to address all warnings, so adding these debug-only checks seems like a good tradeoff to me.