remove unused logback dependency

[elharo]

No description provided.

[cstamas]

Why is there warning in tests about lack of slf4j backend? Which slf4j backend is used now that logback is removed?

[olamy]

we need at least one slf4j impl: https://github.com/apache/maven-build-cache-extension/actions/runs/5543308023/job/15013606896?pr=88#step:6:2248

[gnodet]

we need at least one slf4j impl: https://github.com/apache/maven-build-cache-extension/actions/runs/5543308023/job/15013606896?pr=88#step:6:2248

Agreed, @elharo what's the purpose of this PR ? the dependency is actually used during tests

looks like we need it

[elharo]

1. It cures an unused dependency warning.
2. More importantly it removes yet another idiosyncratic dependency not used anywhere else in maven tossed into the classpath, likely without much thought, some years ago.

https://jlbp.dev/JLBP-1

Logging in Java is a vastly over-engineered mess, and we have only ourselves to blame for it. :-(

[elharo]

The warning is an annoying bug in slf4j. We don't need a logging backend in tests. For that matter, we don't need logging in tests. Logging in tests is a bug. Only failing tests should generate output, and that output should come from assertion failures. Passing tests should generate no output. Otherwise the actual issues get lost in a sea of irrelevant log junk.

[olamy]

@elharo -1
Please revert your commit as it's used dependency.
It's used only in test and very helpful when you need to debug failing tests locally or even in CI

[michael-o]

Slf4j simple is enough here...

[olamy]

Slf4j simple is enough here...

Agree but we need one. I had to debug some issues here in the past with docker on GHA and it was really helpful

[michael-o]

Slf4j simple is enough here...

Agree but we need one. I had to debug some issues here in the past with docker on GHA and it was really helpful

That's the purpose of logging...post-mortem assistance 😁

[HannesWell]

Slf4j simple is enough here...

Agree but we need one. I had to debug some issues here in the past with docker on GHA and it was really helpful

You can configure SLF4J-simple via system properties:
https://www.slf4j.org/api/org/slf4j/simple/SimpleLogger.html
If you want, you can turn it off by default and activate it if needed through corresponding configuration or CLI arguments.

[olamy]

Slf4j simple is enough here...

Agree but we need one. I had to debug some issues here in the past with docker on GHA and it was really helpful

You can configure SLF4J-simple via system properties: https://www.slf4j.org/api/org/slf4j/simple/SimpleLogger.html If you want, you can turn it off by default an activate it if needed through corresponding configuration or CLI arguments.

@HannesWell we perfectly know this but here this PR removes every possible logging solution. As it has been said slf4j-simple is perfectly fine. But WE NEED ONE that's the point.
And not simply remove everything without any replacement!

[olamy]

logging capability restored here 3b318bb

[elharo]

In the short term I'm OK with slf4j-simple but logback should never have been added. It's needless dependency bloat, and every extra dependency is a security risk.

In the long term, nothing beyond the JDK is needed, nor has been since Java 1.4.

I don't think it's a good idea to commit changes without going through PRs and code review. I would have approved 3b318bb if it had been sent through the usual process, but the process should be followed. Ideally the repo should be configured to prevent accidental and deliberate pushes to master. I know I've fumble fingered that one more than once by editing on master when I thought I was on a branch,.

[olamy]

@elharo did you really follow any process here? Some of us added some comments/concerns but you have still merged it.
So I'm trying to figure out your comment about the following process.

[elharo]

Absolutely. I sent a PR. People responded. No one disapproved the PR by sending a "Request changes" review. I responded to the comments, received an approval from another committer, and merged. That is exactly how it is supposed to go.

What is not supposed to happen is anyone committing anything directly to master and then pushing the branch with no approvals or review. 3b318bb could and should have been sent as a PR to be reviewed, and then merged after review and approval like any other change.