Skip to main content

Why Distributed Revision Control increases Agility

I recently did a talk about practices we have introduced to make us more agile.

At the end of the presentation, I mentioned distributed revision control as one of the next practices we want to do. Now how on earth does using Git or Mercurial increase our agility?

The short answer is: Feature Branches.

The long answer is...

We have a defined agile as among other things, this value: Ship working software.

This value implies that we keep the mainline of the code (known to many as the stable branch) in a working state. A good way to keep it that way is by making sure that only safe changes make it into the mainline. If you've got developers continuously working on the mainline to create new features piece by piece, chances are the code is in a broken state every now and then.

In order to remedy this, lots of companies (including ours) have decided that the team should maintain two versions:

  • one experimental, where new features are developed, and 
  • one stable, which should always be in a working state. 

This is a bit of an expensive practice, because you end up with long rounds of verification before the experimental branch can be released, replacing the stable version. You also have to continuously merge back bugfixes from the stable branch to the experimental branch.

So, the idea of feature branches is to keep new development out of the way until it is ready to go out in the mainline. When the feature is complete, you test it well. You then merge the feature into the mainline, do some quick verification that it is still in a working state, and then you release. Much smoother (agile) than releasing an entire experimental branch.

So feature branches are agile. Why don't we do more of those?


The problem is Subversion, or whatever other centralized VCS you're stuck with. They do not allow easy/cheap merging between branches (and I don't want to hear about that half-assed merge-tracking svn-properties hack), and therefore feature branches are practically impossible.
.
You might argue that Subversion users can  keeping their changes locally in the workspace, but there's a major constraint: They can't collaborate on their feature. This means missing out on a lot of co-operation, creative process, ideas, code-review and refactoring before the feature ends up in the mainline. They could operate with sending patches around, but at some point it's time to ask how much hassle you have to take before you replace your malfunctioning VCS with a proper one.

At this point, I would like to outsource the illustration of how you can do feature branches in practice to the very fine illustration from Vincent Driessen. He has really nailed it, I think. In his model, what I've called mainline, or stable, he calls "master", and what I've called experimental, he calls "develop". The release branches also serve as nice points for setting up Continuous Integration (cause you probably don't want to set up CI for every single feature-branch).

So, summary: 


Distributed Revision Control allows you to do feature branches, keep the software in a working state, and ship one feature at a time. In other words, more agile.
Labels:

Comments

  1. fhd25/8/10 15:40

    I thought about this a lot lately and I agree with you: Feature branches are the way to go and they are a PITA in Subversion. Because of that, we do them only for large changes, and it always costs a couple of man hours to set them up and merge them.

    I assume you're still into Git+SVN, do you use feature branches? Does that work well?

    ReplyDelete
  2. tfnico25/8/10 15:49

    Thanks for your comment, Felix.

    We're still on Git+SVN, and I'm afraid that in being compatible with SVN, you are constrained to do branching just like Subversion does it, so there is very little branching possible. You always have to rebase and linearize before the Git commits go back into Subversion, oblivious that they were ever made in a branch.

    That being said, if some of my colleagues are willing to work with me over Git, we can keep our work in a Git branch, before it's finally time to rebase back. I haven't actually done this yet, but I'm pretty certain it will be fine.

    ReplyDelete
  3. fhd25/8/10 17:22

    Sounds good. Our feature branches are usually used by only one person, so that work flow would be fine.

    ReplyDelete
  4. tfnico8/9/10 08:24

    I'm not the first person to conclude this, of course. Jeffrey Palermo did so all the way back in January.

    ReplyDelete
  5. tfnico8/9/10 08:34

    Also here, especially David Arve's paper on the subject.

    ReplyDelete
  6. tfnico8/9/10 09:23

    Digging further through history, I've also found

    ReinH's Git Workflow for Agile Teams which again is based on Josh Susser's Agile git and the story branch pattern.

    ReplyDelete

Post a Comment

Popular posts from this blog

Open source CMS evaluations

I have now seen three more or less serious open source CMS reviews. First guy to hit the field was Matt Raible ( 1 2 3 4 ), ending up with Drupal , Joomla , Magnolia , OpenCms and MeshCMS being runner-ups. Then there is OpenAdvantage that tries out a handful ( Drupal , Exponent CMS , Lenya , Mambo , and Silva ), including Plone which they use for their own site (funny/annoying that the entire site has no RSS-feeds, nor is it possible to comment on the articles), following Matt's approach by exluding many CMS that seem not to fit the criteria. It is somewhat strange that OpenAdvantage cuts away Magnolia because it "Requires J2EE server; difficult to install and configure; more of a framework than CMS", and proceed to include Apache Lenya in the full evaluation. Magnolia does not require a J2EE server. It runs on Tomcat just like Lenya does (maybe it's an idea to bundle Magnolia with Jetty to make it seem more lightweight). I'm still sure that OpenAdvant...
6 comments
Read more

Git tools for keeping patches on top of moving upstreams

At work, we maintain patches for some pretty large open source repositories that regularly release new versions, forcing us to update our patches to match. So far, we've been using basic Git operations to transplant our modifications from one major version of the upstream to the next. Every time we make such a transplant, we simply squash together the modifications we made in the previous version, and land it as one big commit into the next version. Those who are used to very stringent keeping of Git history may wrinkle their nose at this, but it is a pragmatic choice. Maintaining modifications on top of the rapidly changing upstream is a lot of work, and so far we haven't had the opportunity to figure out a more clever way to do it. Nor have we really suffered any consequences of not having an easy to read history of our modifications - it's a relatively small amount of patches, after all. With a recent boost in team size, we may have that opportunity. Also the need for be...
1 comment
Read more

The Git Users Mailing List

A year ago or so, I came across the Git-user mailing list (aka. "Git for human beings"). Over the year, I grew a little addicted to helping people out with their Git problems. When the new git-scm.com webpage launched , and the link to the mailing list had disappeared, I was quick to ask them to add it again . I think this mailing list fills an important hole in the Git community between: The Git developer mailing list git@vger.kernel.org  - which I find to be a bit too hard-core and scary for Git newbies. Besides, the Majordomo mailing list system is pretty archaic, and I personally can't stand browsing or searching in the Gmane archives. The IRC channel #git on Freenode, which is a bit out-of-reach for people who never experienced the glory days of IRC. Furthermore, when the channel is busy, it's a big pain to follow any discussion. StackOverflow questions tagged git , these come pretty close, but it's a bit hard to keep an overview of what questio...
Post a Comment
Read more

Git-SVN Mirror without the annoying update-ref

This post is part of  a series on Git and Subversion . To see all the related posts, screencasts and other resources, please  click here .  So no sooner than I had done my git-svn presentation at JavaZone , I got word of a slightly different Git-SVN mirror setup that makes it a bit easier to work with: In short, my old recipe includes an annoying git update-ref step to keep the git-svn remote reference up to date with the central bare git repo. This new recipe avoids this, so we can simply use git svn dcommit   directly. So, longer version, with the details. My original recipe is laid out in five steps: Clone a fresh Git repo from Subversion. This will be our  fetching repo. Set up a  bare repo. Configure pushing from the fetching repo to bare repo In the shoes of a developer, clone the repo Set up an SVN remote in the developer's repo In the new approach, we redefine those last two steps: (See the original post for how to do the fir...
17 comments
Read more

Git Stash Blooper (Could not restore untracked files from stash)

The other day I accidentally did a git stash -a , which means it stashes *everything*, including ignored output files (target, build, classes, etc). Ooooops.. What I meant to do was git stash -u , meaning stash modifications plus untracked new files. Anyhows, I ended up with a big fat stash I couldn't get back out. Each time I tried, I got something like this: .../target/temp/dozer.jar already exists, no checkout .../target/temp/core.jar already exists, no checkout .../target/temp/joda-time.jar already exists, no checkout .../target/foo.war already exists, no checkout Could not restore untracked files from stash No matter how I tried checking out different revisions (like the one where I actually made the stash), or using --force, I got the same error. Now these were one of those "keep cool for a second, there's a git way to fix this"situation. I figured: A stash is basically a commit. If we look at my recent commits using   git log --graph --...
12 comments
Read more