[
https://issues.apache.org/jira/browse/MRESOLVER-442?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17790540#comment-17790540
]
ASF GitHub Bot commented on MRESOLVER-442:
------------------------------------------
rmannibucau commented on PR #381:
URL: https://github.com/apache/maven-resolver/pull/381#issuecomment-1829749022
@cstamas these are two ok things:
* enforcer config, not a big deal IMHO and anyway enforcer is doomed by mjar
- and once again 2 jars solve it without issues
* IoC/scanners must ignore classes they can't run so it is fine IMHO - and
once again 2 jars solve it without issues since you will not scan the java 11
one
@struberg literally means that `if some dependency has a *NOT* a
multi-release-jar but has additional support for newer java versions, then this
is also perfectly fine as long as it can execute on the requested jvm
version.`, no idea why it wouldn't, this is how most relocaed packages had been
handled since 10+ years without issues.
so overall multi jar release is *always* pointless and it is better to stay
away from it due to the downsides mentionned earlier.
I have really a hard time to see why jumping on a feature which breaks the
library nature of the project is good since technically there is nothing which
can explain it.
> New JDK transport JAR mixes classes with different bytecode
> -----------------------------------------------------------
>
> Key: MRESOLVER-442
> URL: https://issues.apache.org/jira/browse/MRESOLVER-442
> Project: Maven Resolver
> Issue Type: Task
> Components: Resolver
> Reporter: Tamas Cservenak
> Assignee: Tamas Cservenak
> Priority: Major
> Fix For: 2.0.0, 2.0.0-alpha-3
>
>
> And hence is unusable in projects that use Enforcer to enforce highest
> allowed bytecode (like Maven is).
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
