Red Hat Summit 2019 — BPM

At Summit 2018 one of the main focuses in the RH Process Automation Manager presentation was personas. They came up with several personas or user profiles, that were believed to be the main users of PAM, especially Business Central.

I’ve always wondered where they came up with these personas. A marketing study would have been wise. I think it would have showed that while many enterprises intend to have business users develop and maintain processes and rules, it almost always ends up being an IT function.

End users like case managers really need a custom interface built for their specific operation. The Business Central experience isn’t really optimal for that.

At Summit 2019 I was gratified to see that jBPM is headed in a direction that will optimize the experience for the developer. I saw demos of building process and business rule apps using Visual Studio Code using the Quarkus framework. Although not quite ready for general use, they have a stand-alone bpmn editor that is a plugin to VSC. That along with the Quarkus will make jBPM development a real joy.

Now we can interact with a standard git repo and host it on github or elsewhere and not be held back by the limited support in Business Central. The coding environment on Business Central was never a passable experience even for writing short scripts. Using VSC or any other IDE is so much better. Yes, there is a BPMN plugin for eclipse but that seems to have fallen behind the capabilities of the process engine and doesn’t seem to be under active development anymore.

Developing with Quarkus along with the VSC plugin is going to be a great experience. If you open a terminal on your project and run mvn compile quarkus:dev, it will continuously compile your project as the files change (even bpmn and drl) and let you interact with it as you continue development. There’s no manual build, deploy cycle taking your time.

If you are developing a web app, you just refresh the page and try out your new code. You can also run tests at any time. No need to do the build. If you are building for OpenShift or Kubernetes, you can make use of the new Operators to automatically deploy to the platform when the code is checked in.

It’s a big step forward since last year. The jBPM developers have done a great job.