abayer what is the situation with old parallel syntax? When we convert from groovy to JSON will we get a parallel stage block back for the old parallel syntax?

My worry is that when we introduce config at the parallel level in the editor that we will run into serialization issues.

You will not get a parallel stage block back - you'll get the old parallel syntax from the old parallel syntax. No magic.

kzantow abayer hmm how should we handle the case where a user wants to set an agent for a parallel block but their backing Jenkinsfile uses the old style parallel?

Well, I suppose the editor could read old-style parallel but output new-style parallel regardless of whether the config changed? I'm wary of making Declarative itself translate from old-style to new-style when going to JSON, but it might be more reasonable to do that in the editor...

cliffmeyers we'd like you to pick up some more editor work. You will need to organise a handover from kzantow

And some examples: https://github.com/jenkinsci/pipeline-model-definition-plugin/pull/152/files#diff-13c773a7c988e725755ddaa5e6d910a2

Ive added a link to the backend PR>

cliffmeyers can this be your next task?

jamesdumay cliffmeyers I spoke with Andrew about this briefly today as he's eager to get the new parallel syntax in. The changes for this ticket should be fairly trivial and I could probably get it knocked out in a few hours, it only has to do with modifying the declarative JSON conversion functions to a) support the new parallel model and b) always output the new parallel model instead of the old one.

kzantow I've got easily another day left on JENKINS-42791 so if you want to grab this one, by all means go for it. I can still lend a hand w/ the Git public key credential stuff once that one is done - which hopefully helps you out a bit.

cliffmeyers yes, I have just a bit of testing on the git backend and I'll pick this up while waiting on reviews and such; a good switch regarding the public key display, it probably is blocked a bit on what you're working on, anyway, since it's directly related to the same area

Is this one really different to https://issues.jenkins-ci.org/browse/JENKINS-41960 - sounds more and more like a duplicate. 

Seems that way to me, yes.

michaelneale this is effectively the same as JENKINS-41960 I'll enable the same stage editing as part of this ticket.