And now I was able to reproduce this:

https://github.com/mfriedenhagen/plugin-classpath-maven-3-and-4 
<https://github.com/mfriedenhagen/plugin-classpath-maven-3-and-4>

mvn clean verify succeeds with 3.6.3 and breaks with 
619973b91cff5af7b5695bda11761a982a159349 ff.



> Am 23.02.2021 um 14:42 schrieb Mirko Friedenhagen <mfriedenha...@gmx.de>:
> 
> Not that deep in bisecting as of now. I wrote a small script `bisect`:
> 
> ```
> #!/bin/bash -e -p
> mvn -q -B -s ~/.m2/settings-oss.xml clean package -Dmaven.test.skip=true 
> -Denforcer.skip=true -Drat.skip=true
> bin_tar_gz=apache-maven/target/apache-maven-[3,4]*-bin.tar.gz
> test -f $bin_tar_gz
> gunzip -dc $bin_tar_gz | (cd /tmp/ && rm -rf apache-maven-*-SNAPSHOT && tar x)
> cd $HOME/ghq/gitlab.example.com/group/failing-project
> /tmp/apache-maven-*-SNAPSHOT/bin/mvn -B -V -q clean verify -DskipTests
> ```
> 
> Running:
> ```
> $ git bisect start 84a1ff2d5cb7bf49fa3171dd92f7ac088e837185 maven-3.6.3
> $ git bisect run ./bisect
> ```
> 
> led to:
> 
> ```
> $ git bisect log
> # bad: [84a1ff2d5cb7bf49fa3171dd92f7ac088e837185] [MNG-7063] Add unittest to 
> ensure rereading a pom won't return the cached Model
> …
> # first bad commit: [619973b91cff5af7b5695bda11761a982a159349] [MNG-7063] 
> Infinite loop using Shade plugin and JUnit 5 dependency
> ```
> 
> Best Regards
> Mirko
> 
> 
>> Am 23.02.2021 um 10:21 schrieb Michael Osipov <micha...@apache.org>:
>> 
>> Am 2021-02-23 um 10:16 schrieb Mirko Friedenhagen:
>>> Hello Michal, hello Robert,
>>> Thanks for the answers.
>>> I am trying to reproduce this with a smaller project, but it is quite hard. 
>>> The problem occurs with an inhouse project where we use inhouse plugins 
>>> etc. We manage the lifecycle of standard packagings as well in the the 
>>> extension. I wrote a FOSS derivate (not widely use) and will try to give a 
>>> small example.
>>> Robert, did you mean to bisect the maven-core?
>> 
>> Yes, he did. This is what I would do.
>> 
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org
>> For additional commands, e-mail: dev-h...@maven.apache.org
>> 
> 
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org
> For additional commands, e-mail: dev-h...@maven.apache.org
> 

Reply via email to