From https://github.com/jenkinsci/workflow-cps-plugin/commit/f3958997d201d39a952c6572eb623a8730b822d7, it looks like the Global Variable section from the old block.jelly wasn't copied over to the new index.jelly - jglick, was that a deliberate omission?

I tried copy-pasting it into block.jelly and it didn't work - it gets the dropdown list of the GlobalVariable that are available but doesn't render the help.

Whoops, when I moved it around it worked. PR incoming.

PR up - https://github.com/jenkinsci/workflow-cps-plugin/pull/17

Yes it was a deliberate omission.

Latest PR - https://github.com/jenkinsci/workflow-cps-plugin/pull/18

In my experience of Groovy the distinction between different ways to getting at things is pretty vague. I understand that a properties, methods, fields, and instances are diffent, but from a usage standpoint, the language seems to be designed to minimize one's awareness of those differences. From an end-user/pipeline-writer perspective, treating the global variables differently than we treat method is more confusing that helpful.

I'll second what bitwiseman is saying, as a Pipeline user, the distinction between different flavors of API reference documentation is not really relevant to me. With the docker variable, for example, there are methods hanging off of that which I think of, and work with, like other "steps".

The docker variable is a different case, and to deal with that properly really requires a separate API: JENKINS-32731.

Actually got merged a while back.