Devin Nusbaum

@chickenkiller If you still have your Jenkins logs from earlier, do you see a message like "Preventing /org/csanchez/jenkins/plugins/kubernetes/pipeline/KubernetesDeclarativeAgentScript.groovy from being loaded without sandbox protection in . To allow access to...

IDK. While testing I sometimes saw two instances of `WithScriptAllowlist` when trying to print `WithScriptAllowlist.scriptUrls` from the script console, but this only ever happened when loading plugins dynamically. > Could...

> @dwnusbaum was most recently reviewing seriously No, my recent review was just to try to help resolve https://github.com/jenkinsci/workflow-cps-plugin/pull/577#issuecomment-1355502611 and a bad merge related to https://github.com/jenkinsci/workflow-cps-plugin/pull/612, which is something I...

We just ran into a similar issue in the Jenkins project in https://github.com/jenkinsci/workflow-cps-plugin/pull/901, but we are using JDK 17, so it was a fatal error rather than a log warning....

@timja Yes sorry I'm just running the OSS BOM and another proprietary PCT suite to see if it's worth trying to adjust `getContext` to only be called when a stage...

I pushed https://github.com/jenkinsci/pipeline-model-definition-plugin/pull/700/commits/bd144259a8a92027109360538429b4f9e7813e51 to try to reduce the number of impacted tests and will rerun the PCT once a new incremental build is available. If nothing else it at least...

I triggered another PCT run (just a CloudBees internal one this time) and will check the results tomorrow.

@jglick it's in the "Issue" `` in the description.

@cyrille-leclerc `GraphListenerAdapterToPipelineListener` should probably not extend `GraphListener.Synchronous`, just `GraphListener`, see https://github.com/jenkinsci/workflow-api-plugin/blob/ca5fddb3fcebcc6acf823c80600ec7ab9deb0d05/src/main/java/org/jenkinsci/plugins/workflow/flow/GraphListener.java#L48. Or perhaps you could modify it to only perform a small subset of the current logic synchronously if necessary...

The serialization compatibility mistake in #223 (changing the type of a serialized field from `Calendar` to `Map`) that was reported in #323, #324, and again in this issue, seems to...