The usual way to maintain a binding is to:
- Clone the openhab/openhab repository
- Create a new branch in your clone for your new binding
- Add your code to your branch
- Test, build, refine, write a wiki page to document its purpose and use, find testers, get feedback, etc.
- Make sure the JAR works correctly in openHAB 2, even if you normally use openHAB 1.x.
- Once it’s ready for inclusion in the openHAB repository, submit a pull request from your branch.
- An openHAB maintainer will review your binding and request (hopefully minor) changes. (This is a resource-constrained process at present.) You can keep making commits to your branch in your clone and they will appear in the pull request.
- When it’s signed off on, you will likely have to squash multiple commits into one, so a single commit represents all your changes to add the binding to openHAB.
- The openHAB maintainer will merge your pull request into the openhab/openhab repository.
- You add your documentation wiki page to the sidebar list of bindings.
- As any new bugs or enhancements come up, you (or some other contributor) will submit new pull requests based on new branches created for the purpose of individual bugs or enhancements.
Soon these steps will be different, but this is basically how it’s been done in recent history.