Add a mechanism to suppress stack trace.

[jeffret-b]

Add a mechanism to control whether HttpResponses.error() shows the error details, particularly the stack trace. By default it doesn't show the stack trace, but the interface can be enabled to show it.

[jvz]

Seems like a reasonable approach.

[jeffret-b]

@Wadeck @daniel-beck

[jeffret-b]

Let's see if incrementals are working yet.

[jeffret-b]

Try kicking off incrementals again.

[daniel-beck]

I wonder whether it would be nicer to use a similar pattern as I used for the RequirePOST error page that lets the application render a view: https://github.com/stapler/stapler/blob/7d5bf955f3f0d174e760c522359ad8321c946f97/core/src/main/java/org/kohsuke/stapler/interceptor/RequirePOST.java#L40-L63

Not sure we're in the correct level here to do this, though. That way, even Stapler's error pages (and thereby the error pages served by the application through HttpResponses#error) would be nicer looking.

As this extends the API of Stapler, bringing this idea up now instead of considering this just a future enhancement. Since nobody else uses Stapler, it would be easy enough to defer though.

[jeffret-b]

I adopted Daniel's suggestion of tying the implementation back to Jenkins again, this time using the oops page. I was concerned that we would have things in the proper space and stack but it seems to work fine.

Please review @jvz @daniel-beck @Wadeck

[jeffret-b]

It looks like this is ready to go. @jglick or @daniel-beck, who can help get this merged in and released?

[daniel-beck]

Looks like Stapler is (now?) released to repo.j.o, and nobody has permissions to do this, which basically means Oliver Gondza and the Artifactory admins (KK, Tyler, Wadeck, and me) can do it. Or we use some of the security process to deploy into a private repo and them copy it over. 😆

(I removed @jglick's previous universal upload permission a while back.)

[jglick]

IIUC this is waiting for jenkins-infra/repository-permissions-updater#1420

[daniel-beck]

@jglick Looks like the parent POM results in deployment to OSSRH after all? If so, could you release Stapler?

[jglick]

https://github.com/stapler/stapler/releases/tag/stapler-parent-1.259

[jeffret-b]

Thanks for releasing that, @jglick!