Before developing a patch for pgAdmin you should always contact the developers on the mailing list pgadmin-hackers @ postgresql . org to discuss your plans. This ensures that others know if you’re fixing a bug and can then avoid duplicating your work, and in the case of large patches, gives the community the chance to discuss and refine your ideas before investing too much time writing code that may later be rejected.

You should always develop patches against a checkout of the source code from the GIT source code repository, and not a release tarball. This ensures that you’re working with the latest code on the branch and makes it easier to generate patches correctly. You can checkout the source code with a command like:

$ git clone git://git.postgresql.org/git/pgadmin4.git

Once you’ve made the changes you wish to make, commit them to a private development branch in your local repository. Then create a patch containing the changes in your development branch against the upstream branch on which your work is based. For example, if your current branch contains your changes, you might run:

$ git diff origin/master > my_cool_feature.diff

to create a patch between your development branch and the public master branch.

You can also create patches directly from the development tree, for example:

$ git diff > my_cool_feature.diff

If you are adding new files, you may need to stage them for commit, and then create your patch against the staging area. If any of the files are binary, for example, images, you will need to use the –binary option:

$ git add file1.py file2.py images/image1.png [...]
$ git diff --cached --binary > my_cool_feature.diff

Once you have your patch, check it thoroughly to ensure it meets the pgAdmin Coding Standards , and review it against the Code Review Notes to minimise the chances of it being rejected. Once you’re happy with your work, mail it as an attachment to the mailing list pgadmin-hackers @ postgresql . org . Please ensure you include a full description of what the patch does, as well as the rationale for any important design decisions.