Skip to main content

"Solving" a failing unit test...

No replies

... by commenting the offending line? Just noticed, that's exactly what
happened to a failing method in JXTitledPaneTest (it had been failing
ever since the api changed ... not pointing fingers )

IMO that's about the worst thing to do. If the code owner (which
admittedly is a very vague term in our setup - but not me in this
concrete case ;-) didn't do anything about it while it failed, how
probable is a fix once it disappears from an publicly viewable failure
into a hidden "TODO" marker?

BTW, we do have an intermediate: unit tests included into a build are
those postfixed with "Test", unit tests not included into a build are
those postfixed with "Issues" (actually everything except "Test") - so
we can keep known issues alive without burying them ... Personally, I
love the pleasure to promote a formerly failing method from "Issue" to
passing method in the corresponding "Test".


To unsubscribe, e-mail:
For additional commands, e-mail: