maven-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Dan Fabulich (JIRA)" <>
Subject [jira] Closed: (SUREFIRE-464) Default for failIfNoTests should be false
Date Tue, 18 Mar 2008 01:42:46 GMT


Dan Fabulich closed SUREFIRE-464.

    Resolution: Won't Fix

I think you've made some incorrect assumptions about the way failIfNoTests works.

In Surefire 2.4 we added some sweeter sugar to skip testing: -DskipTests.  (You don't even
have to say "=true", just "-DskipTests" will do the trick.)

As a result, there's no reason to use -Dtest=nonexistent, -Dtest=0, -Dtest=false or whatever,
so we dropped its sugary use, in favor of adding better error protection for when you're trying
to use it the normal way but you make a typo, e.g. "-Dtest=MyTset".

-Dtest will cause failIfNoTests to become true in a way you can override, but the default
value for failIfNoTests is actually false (because the default -Dtest is undefined).  It is
already the case that the build will pass if you simply build a project that has no tests
(without using -Dtest).

In summary, this bug is wrong, partly because failIfNoTests is already false by default (though
it's slightly more complex than that), and partly because -Dtest=nonexistent *should* cause
an error.

> Default for failIfNoTests should be false
> -----------------------------------------
>                 Key: SUREFIRE-464
>                 URL:
>             Project: Maven Surefire
>          Issue Type: Bug
>          Components: plugin
>    Affects Versions: 2.4, 2.4.1, 2.4.2
>            Reporter: Dain Sundstrom
> In maven the surefire plugin executes by default when ever I build, and since the default
for failIfNoTests is now true all of my modules must have at least one test or I must manually
configure the plugin to not fail.  This is a minor annoyance for single module builds but
is a major pain when you have lots of modules.  In my case it is extremely annoying since
we have example modules we distribute that don't extend or parent pom (so users don't have
to pull down our whole code base to get an example project).
> This also means that the syntactic sugar "-Dtest=false" to turn off all testing no longer
> At the end of the day, I don't understand why not having tests is a build failure and
not just a warning.

This message is automatically generated by JIRA.
If you think it was sent incorrectly contact one of the administrators:
For more information on JIRA, see:


View raw message