So what I have is data in an application database, and I’ve got flowable sitting as an independent system. There’s a flowable workflow that monitors an approval process for changes made to the data. I have a UI screen that required a query to combine both the flowable workflow state (and some process instance variables) with the application data. I “solved” this problem by creating a database view on the flowable database that exposed some of the flowable process instance and task history tables, then on my application side I joined this view to my application data tables (via the process ID, which I used as the primary linkage between everything).
I did it this way because if I used the java APIs, I would have ended up with a double loop (first getting all the application data and then looping the flowable data via the APIs, or first using the flowable APIs and then looping all the application data).
My question is: Is this the best solution to this problem, or am i setting myself up for headaches in the future if the table structures are altered? Right now I’m on 6.4, but I could see the tables getting tweaked in the future, and that could come with risks.