See #12544 for the fixes.

@okonomiyaki3000 Tested like written in instructions, see with or without Patch Errors or not. So don't know what exact to look for.

@franz-wohlkoenig It should not be difficult to find errors. Any page of the administrator should at least produce: Uncaught SyntaxError: Octal literals are not allowed in strict mode.

I have tested this item ? unsuccessfully on 1914619

Tested in J3.7.0-alpha2-nightly on different Isis-Menues, got no Uncaught SyntaxError: Octal literals are not allowed in strict mode. With and without PR got Warnings like getAttributeNode() shouldnt used. Use getAttribute().

Excellent. I think 'getAttributeNode()' has nothing to do with this file though. Where does your js console say it appears?

The 'octal literals' bug is actually caused by a line like this:

var cls = this.className.replace(/^.(chzn-color[a-z0-9-_]*)$.*/, '\1');

You will notice that there are no octal literals here. The problem is the replacement string. It should be '$1', not '\1'. This line still seems to be in 3.7 so I'm not sure why your browser is letting it go. May I ask which browsers you've tested this in?

Anyway, keep in mind that this PR is not meant to fix anything and is not expected to be merged. It is here as a reference to expose problems which normally go unnoticed because strict mode is not used. The goal of this patch is to make things break. The fixes are in #12544. That is the PR that should actually be merged.

@okonomiyaki3000 Used Firefox 50.1.0, thanks for explanation of this PR. Will test #12544

I have tested this item ✅ successfully on 2392574

Now I'm getting Uncaught SyntaxError: Octal escape sequences are not allowed in strict mode.

@okonomiyaki3000 could you please look into: #21233 Thanks

@Tchangue Yes, this PR has errors. That's actually the only point of it. It is not meant to be merged. It meant to expose the errors by invoking strict mode. The errors are actually fixed in #12544.

@FPerisa You got errors. Which means the patch works because it's supposed to expose errors. But I think you're supposed to mark it as failed because we actually do not want errors. Yes, it seems to make no sense. The problem is that the errors are there already, they're just not reported because we're not in strict mode. So in fact, this PR should be Failed and the current Joomla code should also be Failed. #12544 is correct.

I have tested this item ✅ successfully on 2392574

I think this issue can be closed together with #12544 because #12543 was only a patch to allow the test of #12544

Ready to Commit after two successful tests.

@franz-wohlkoenig Do not commit.

@kneisel you can't have tested this patch successfully. It intentionally has errors.

@okonomiyaki3000 you mean set RTC back on Pending?

sorry, this is closed.

This is fine, it was closed without merging. As it should be. We're all fine here now.