-
Notifications
You must be signed in to change notification settings - Fork 213
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Regression mvnd 1.0-m6 vs. mvn 3.9.1 #841
Comments
pretty ping? |
Quick update: I built the current trunk from scratch with
and tried both the m39 and m40 variants. Same result. So this is a problem that still exists in the current version and it is not confined specifically to maven 3.9 or maven 4.0 |
This also fails only with the |
@hgschmie thanks for tracking this down and filing the issue. We've been hitting this as well. I can confirm that putting |
So this problem is caused by a combination of two factors:
What happens is that when the first module is built, the enforcer loads the rules, including the |
My analysis was wrong. The problem seems to come from the MavenSession injected into plexus components for the enforcer realm, which does not point to the correct project. |
The fix is available at takari/takari-smart-builder#25 |
how to reproduce:
Same command with mvn works fine.
Given the way our project sets module names (we use per-module profile activation based on the presence and absence of files in the directories), the most likely culprit is some sort of caching of profile activation. The correct list of profile activations would be:
The
in-parent
profile provides a module name. Running the same command withmvnd
yields no output (!).output of
mvnd -X -pl :jdbi3-core -am clean install
logfile-mvnd.txt
output of
mvn -X -pl :jdbi3-core -am clean install
logfile-mvn.txt
The text was updated successfully, but these errors were encountered: