[
https://issues.apache.org/jira/browse/LOG4J2-3614?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17614075#comment-17614075
]
Volkan Yazici commented on LOG4J2-3614:
---------------------------------------
LOG4J2-3075 was a bug fix, not a feature change. Hence nothing new was
introduced there, except fixing the formatting of the nanoseconds. To be
precise, caching on the incorrect key was the problem.
Timestamp resolver still behaves as documented. Mind showing an example where
you suspect an inconsistency, please?
> Update documentation after LOG4J2-3075
> --------------------------------------
>
> Key: LOG4J2-3614
> URL: https://issues.apache.org/jira/browse/LOG4J2-3614
> Project: Log4j 2
> Issue Type: Bug
> Components: Documentation
> Affects Versions: 2.17.2
> Reporter: Strainu
> Assignee: Volkan Yazici
> Priority: Minor
>
> After LOG4J2-3075 was implemented, the JsonTemplateLayout timestamp resolver
> documentation should be updated to reflect the new feature.
> It's especially important to do so since the second fragment identifier used
> here ('S') is different from the one described in, e.g.
> [https://logging.apache.org/log4j/2.x/manual/layouts.html] (which is 'n'). I
> don't have enough experience with Log4j to determine if this inconsistency is
> a bug or not, but an entry in the JsonTemplateLayout documentation would
> help.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
