Hi,
It’s great that you are interested in running Camunda on a new database technology and considering a contribution. I want to expand a bit on Niall’s repsonse: Adding support for a new database technology to the product is a pretty costly operation for us. In terms of code it may be not too complex, however the lesser visible efforts are:
- We need to constantly cover this technology in our CI to avoid regressions
- We need to be confident in supporting our enterprise users with this technology
- When building new features in the future, we are more restricted in the SQL statements that we can write (as we are mostly tied to the minimal intersection of features that all our supported databases provide)
In that sense, officially supporting a new database technology is not something that we do lightly and we may reject a community contribution to the core codebase. This depends a lot on the overall demand for the technology and Camunda.
That said, this is not necessarily the end of the story:
- @langfr maintains his own fork of Camunda with Informix support: GitHub - langfr/camunda-bpm-platform: Flexible framework for workflow and process automation (see master branch). You consider do something similar. It’s also a great example to see what parts of the codebase you may have to touch.
- Instead of a fork, you could also try to build this as a community extension. If the process engine misses extension points to achieve this (which I somewhat expect), then we could talk about contributing these extension points so that it becomes easier to plug in your own database support
Cheers,
Thorben