IMP includes several pieces of code that are actually maintained in separate git repositories on GitHub. These include the RMF library and the IMP::pmi module. There are two main ways to achieve this with git: git submodule and git subtree. Modern versions of IMP use git subtree (we previously used submodule; see the issue on GitHub for a discussion). subtree is a little harder for developers to work with than submodule, but is much easier for end users (no need to run any setup scripts after getting IMP, forks of the IMP repository work, and GitHub's "Download ZIP" feature works).

This blog post covers several useful commands for working with git subtree. However, git subtree commands are hard to use correctly can be very slow, so it's generally easier to use this procedure when working on PMI:

• Build IMP as per usual, say in ~/imp/build/.
• Don't touch the files in ~/imp/modules/pmi; instead git clone the PMI repository into a new directory, separate from IMP, say ~/pmi/.

• Make a new directory for the PMI build, e.g. ~/pmi/build, then do an out-of-tree build of PMI with:

  cmake .. -G Ninja -DIMP_DIR=~/imp/build/ && ninja

• Use regular git commands git add, git commit, git push etc. to record changes to the PMI repository. Test changes using ctest or the setup_environment.sh script in ~/pmi/build.

• To incorporate changes from the PMI repository into IMP, first push them to the PMI repository, then use the utility script (in the IMP clone)

  tools/git/update-pmi.sh

  • This will squash all of the PMI changes into a single IMP commit, which can then be git pushed into the IMP repository in the normal way.