Documentation needed: How to work with multiple (patch) orgs and branches in single IC2 project

Issue #861 new
Former user created an issue

I am a developer from a Salesforce ISV with multiple managed packaged. I recently switched from Eclipse and can't get this use case to work with IDEA and IC2.

In Eclipse I had a project for the packaging org (reflected the trunk of my Github product repo) and n other projects for the different patch orgs of the same product. Each patch project was connected to a Github branch of the product repo.

I had all those projects in a per product Workset in Eclipse so I could easily work in the trunk, compare it to patches, merge code and the like.

After talking to Scott from IC2 I heard that are features to allow that even better but I couldn't get it to run. I think thats to complex to figure it out by clicking around so I would like a detailed documentation or even better a decently detailed video.

Comments (3)

  1. Scott Wells repo owner

    I'll add some quick thoughts here, but I do plan to provide more first-class documentation (likely in the form of HOWTOs) for things like this.

    When working with multiple branches and/or orgs against what is effectively the same code base, you can use a combination of IntelliJ/WebStorm and IC features to achieve what (I think) you want. I'll use Git for the sake of example here, but the same concepts should apply to any supported version control system in IntelliJ/WebStorm.

    Let's assume for a moment that you typically release software out of the master branch. Your developers create transient feature branches for their work, and you may also create more lasting branches for patch releases. Each branch is associated with a distinct organization. For example, master may be associated with a sandbox, production, or packaging org. Each developer feature branch is associated with that developer's personal DE org. Patch release branches are associated with patch orgs.

    In IC, orgs are modeled as connections, so that means that you have multiple configured con,nections. You select a single connection for each IC module in a project, and IC generates an OST for each connection that you use.

    So now let's make this concrete. When working with the master branch, I configure IC for a connection called packaging which has the credentials of the official packaging org. When working on features, I configure IC for a connection called dev which has the credentials of my DE org. When working on a patch release, I configure IC for a connection called patch<version> which has the credentials of the patch org from which I'll release patch <version>.

    Note that IntelliJ/WebStorm and IC config files can be checked into VCS, and there's actually a huge benefit to doing so. You can find details in the FAQ under Can I manage the IDE project files in version control? Which files should or should not be checked in?.

    This is important because when I'm working in master and have configured IC for the packaging connection, I have an OST under IlluminatedCloud/packaging/ Similar with other branches and their specified connections. If I check in the IDE/IC config files and the OST on each branch, as I switch back and forth between branches things update correctly. There's no need to switch connections or generate an OST after switching to a given branch. Note that even if you don't check in OSTs, this will work. You'll just end up with multiple IlluminatedCloud/<connectionName>/ entries, one per-connection.

    IntelliJ/WebStorm makes it very easy to compare the contents of branches. Here's how you do it in Git, but JetBrains' docs also show how to do it in other supported VCSs.

    Similarly, if you'd like to compare your local project metadata with the contents of another org, start a Retrieve Metadata action, select the desired connection in the Connection drop-down, and click Retrieve for Merge. IC will pull the requested metadata from the specified org into a temporary location and open the directory comparison tool against the project metadata for comparison.

    I think I'll take a break and see if this helps to answer your question. We can iterate a bit based on follow-on questions, and once we have a good explanation, I'll move this into the FAQ until I can produce the aforementioned HOWTOs.

  2. Scott Wells repo owner

    Thanks, Robert. I'll take a look and either document how the same process would work in IC or, if that's not possible, think about how to streamline the IC process.

  3. Log in to comment