Hi David,
I've been working my way through this great resource in order to try and
make heads and tails of maven:
http://www.sonatype.com/books/maven-book/reference/public-book.html
Especially taking notes on:
http://www.sonatype.com/books/maven-book/reference/optimizing.html
It seems that there are a lot of inheritance issues with the current
trunk which leads to a lot of duplicated plugins and deps down the
chain, but often with differing versions. The final build basically does
the best job it can at resolving all required versions.
I know it is a huge and dangerous task to make wide ranging pom changes,
and I have just bitten the bullet and made some changes to several poms.
The question is, how happy are you for commiters (I guess I mean me, but
also anyone interested) to try and clean up some of these issues in
trunk? - Or would a pure Java 1.6 branch be the safer option, at least
until wide ranging changes are tested by several devs on varying platforms.
I would like to put something as drastic as this in the top pom (which
overrides the Apache 7 pom), and work down from there:
<plugin>
<groupId>org.apache.maven.plugins</groupId>
<artifactId>maven-compiler-plugin</artifactId>
<version>2.3.2</version>
<configuration>
<source>1.6</source>
<target>1.6</target>
</configuration>
</plugin>
I'd also like to ensure that all poms reference their parents throughout
the project so that it is possible to pull up plugin and dependency
definitions as high as possible.
This cannot be done without builds initially failing until all issues
are resolved - Issues meaning all the ones that arise on 'other' devs
machines when everything seems fine on a local build ;-)
Regards,
Andy.
