Skip to content

Tag: semver

Detecting Breaking Changes

When integrating with other applications and libraries it’s important to detect when APIs are changing in an incompatible way as that might cause downtime for the downstream application or library. SemVer is a popular versioning strategy that can hint about breaking changes by bumping the major version part but as an upstream application developer it can be difficult to detect that a code change is in fact a breaking change.

ApiCompat

Microsoft.DotNet.ApiCompat is a tool built for .NET that can compare two assemblies for API compatibility. It’s built and maintained by the .NET Core team for usage within the Microsoft .NET development teams, but it’s also open source and available to anyone.

Installation

The tool is provided as a NuGet package on the .NET Core team’s NuGet feed, which is not on nuget.org, which most package managers reference by default, but a custom NuGet server hosted in Azure. The URL to the feed needs to be explicitly specified in the project that like to use it.

In the project file add a reference to the NuGet feed under the Project group:

<PropertyGroup>
  <RestoreAdditionalProjectSources>
    https://pkgs.dev.azure.com/dnceng/public/_packaging/dotnet-eng/nuget/v3/index.json;
  </RestoreAdditionalProjectSources>
</PropertyGroup>

Add a reference to the Microsoft.DotNet.ApiCompat package:

<ItemGroup>
  <PackageReference Include="Microsoft.DotNet.ApiCompat" Version="7.0.0-beta.22115.2" PrivateAssets="All" />
</ItemGroup>

The most recent version of the package can be found by browsing the feed.

Usage

The tool can execute as part of the build process and fail the build if the current source code contains changes that is not compatible with a provided contract, an assembly from the latest major release for example.

While it is possible to commit a contract assembly to Git, another approach is to automatically fetch it from source early in the build process. This example uses a NuGet feed as the release source, but it could be an asset in a GitHub release as well or something else.

<Target Name="DownloadLastMajorVersion" BeforeTargets="PreBuildEvent">
  <DownloadFile SourceUrl="https://www.nuget.org/api/v2/package/MyLibrary/2.0.0" DestinationFolder="LastMajorVersionBinary">
    <Output TaskParameter="DownloadedFile" PropertyName="LastMajorVersionNugetPackage" />
  </DownloadFile>
  <Unzip SourceFiles="$(LastMajorVersionNugetPackage)" DestinationFolder="LastMajorVersionBinary" />
</Target>

This will download a NuGet package to the folder LastMajorVersionBinary in the project directory using the DownloadFile task. If the directory doesn’t exist it will be created. If the file already exist and has not been changed, this becomes a no-op.

The Unzip task will unpack the .nupkg file to the same directory as the next step. Same thing here, if the files already exists this is a no-op.

The last step is to instruct ApiCompat to target the unpacked assembly file as the current contract the source code will be compared with. This is done by setting the ResolvedMatchingContract property, which is the only setting required to run the tool.

<ItemGroup>
  <ResolvedMatchingContract Include="LastMajorVersionBinary/lib/$(TargetFramework)/$(AssemblyName).dll" />
</ItemGroup>

The path points to where the assembly file is located in the unpacked NuGet package directory.

Building the project will now download the contract and execute ApiCompat with the default settings. Remember to use proper access modifiers when writing code. internal is by default not treated as part of a public contract1 as the referenced member or type cannot be reached by other assemblies, compared to public.

1 InternalsVisibleToAttribute can make internal member and types accessible by other assemblies and should be used by care.

Handling a Breaking Change

Breaking changes should be avoided as often as possible as they break integrations downstream. This is true for both libraries and applications. An indication of a breaking change should most of the time be identified and mitigated in a non-breaking fashion, like adding another API and deprecating the current one. But at some point the old API needs to be removed, thus a breaking change is introduced which will fail the build.

The property BaselineAllAPICompatError can be used to accept breaking changes. The specification of the breaking changes will be written to a file called ApiCompatBaseline.txt in the root of the project. ApiCompat uses it to ignore these tracked incompatible changes from now on. This property should only be set when a breaking change should be accepted and should result in a new major release.

<PropertyGroup>
  <BaselineAllAPICompatError>true</BaselineAllAPICompatError>
</PropertyGroup>

Once a new build has be executed and the new contract baseline has been established remember to set the property to false or remove it.

The baseline file should be committed to source control where it can be referenced as documentation for the breaking change by for example including it in a BREAKING CHANGE conventional commit message.

Once a new major release has been published, there are two options.

  1. Reference the assembly in the new major release as the contract and delete ApiCompatBaseline.txt.
  2. Do nothing 🙂

As git can visualize diffs between two references, the content of the ApiCompatBaseline.txt will show all the breaking changes between two version tags, which can be quite useful.

Summary

ApiCompat is a great tool for automating detection of breaking changes during development. It avoids introducing incompatible changes and potential headache to downstream consumers, both for applications and libraries.

A complete example of the mentioned changes are available here.

Leave a Comment

Trunk Based Release Versioning – The Simple Way

Some while ago I wrote an article about release versioning using Trunk-Based development where I used GitVersion for calculating the release version. Since then I have experienced multiple issues with that setup up to the point when I had enough and decided to make my own release versioning script, which I simply call Trunk-Based Release Versioning!

There is no complex configuration, it’s just a bash script that can be executed in a git repository. It follows the Trunk-Based Development branching model and identifies previous release versions by looking for SemVer formatted tags. It outputs at what commits the next release starts and ends and includes the version and the commit messages. Bumping version is done by writing commit messages that follows the Conventional Commits specification.

Trunk-Based Development For Smaller Teams

This is the simplest strategy, there is only the default branch. A release contains everything between two release tags.

Scaled Trunk-Based Development

Development is done on short lived development branches. It is possible, and preferable, to continuously release by merging into the default branch. All commits between two release tags make up a release, same as for the previous strategy, but it is also possible to pre-release from the development branch as pictured by the red boxes. A pre-release contains all commits on the branch since last release and when merging it all becomes the new release. Note that previous merge commits from the development branch onto the default branch also are considered release points in order to be able to continuously work from a single branch. This also works if merging the opposite way from the default branch into the development branch in order to for example enforce gated check-ins to the default branch.

For a full example using Github Actions, see this workflow.

That’s it! Keep releasing simple.

Leave a Comment

Semantic Release Notes with Conventional Commits

Recently I wrote about how to use Conventional Commits to calculate the next release version based on semantic versioning. Today I’m going to show how to turn conventional commit messages into release notes using a simple Github Action called Semantic Release Notes Generator that I’ve written.

release-notes-generator

The action is basically just a wrapper for semantic-release/release-notes-generator which is a plugin for semantic-release, an opinionated version management tool. When trying out semantic-release I found it to be too much functionality entangled in a way I couldn’t get to work the way I wanted it to, but I really liked the release notes generator! Being a fan of the Unix philosophy, I decided to wrap it in a neat little action that I could pipe.

Usage

The generator fetches all commit messages between two git references, and feeds them into the release-notes-generator that formats the messages into a nice looking release log.

The semantic release notes generator actually uses a Github Action workflow to test itself (more about that in another post). It uses gitversion to determine next version and pipe that to the generator to generate release notes based on the commits logged since the last release.

Here’s how that could look like:

...
- name: Determine Version
  id: gitversion
  uses: gittools/actions/gitversion/execute@v0
  with:
    useConfigFile: true
    configFilePath: .github/version_config.yml
- name: Determine Release Info
  id: release
  run: |
    from_tag=$(git tag --points-at ${{ steps.gitversion.outputs.versionSourceSha }} | grep -m 1 ^v[0-9]*\.[0-9]*\.[0-9]* | head -1)
    tag=v${{ steps.gitversion.outputs.majorMinorPatch }}

    echo "::set-output name=tag::$tag"
    echo "::set-output name=from_ref_exclusive::$from_ref_exclusive"
- name: Create Tag
  uses: actions/github-script@v3
  with:
    script: |
      github.git.createRef({
        owner: context.repo.owner,
        repo: context.repo.repo,
        ref: "refs/tags/${{ steps.release.outputs.tag }}",
        sha: "${{ steps.gitversion.outputs.sha }}"
      });
- name: Generate Release Notes
  id: release_notes
  uses: fresa/release-notes-generator@v0
  with:
    version: ${{ steps.release.outputs.tag }}
    from_ref_exclusive: ${{ steps.release.outputs.from_ref_exclusive }}
    to_ref_inclusive: ${{ steps.release.outputs.tag }}
- name: Create Release
  uses: softprops/action-gh-release@v1
  with:
    body: ${{ steps.release_notes.outputs.release_notes }}
    tag_name: ${{ steps.release.outputs.tag }}
...

It calculates the next version and finds the last release’s git reference. Then it creates a tag to mark the next release, which also will be used as to_ref_inclusive when generating the release notes. Lastly it creates the Github release using the release notes and the tag created.

If there’s a need to change the release notes, it’s always possible to edit the release in Github either after the release has been published or by releasing it as a draft and manually publishing it after a review.

Summary

release-notes-generator is a great tool for generating release notes when using conventional commits. With Semantic Release Notes Generator you can in an unopiniated way choose how to use release notes within your automated continuous delivery processes!

Leave a Comment

Simple Release Versioning with Trunkbased Development, SemVer and Conventional Commits

I have scripted versioning strategies for Trunk Based Development many times thoughout my career, including using simple Github Actions like Github Tag Action, but the result was never spot on. So I finally gave in overcoming the complexity of tuning GitVersion!

Trunk Based Development

From trunkbaseddevelopment.com:

A source-control branching model, where developers collaborate on code in a single branch called ‘trunk’ *, resist any pressure to create other long-lived development branches by employing documented techniques. They therefore avoid merge hell, do not break the build, and live happily ever after.

A great fit for Continuous Delivery, and very easy to apply!

SemVer

Versioning does not need to be just about incrementing a number to distinguise releases. It can also be semantic; alas SemVer, or short for Semantic Versioning.

From the site:

Given a version number MAJOR.MINOR.PATCH, increment the:

  1. MAJOR version when you make incompatible API changes,
  2. MINOR version when you add functionality in a backwards compatible manner, and
  3. PATCH version when you make backwards compatible bug fixes.

Additional labels for pre-release and build metadata are available as extensions to the MAJOR.MINOR.PATCH format.

This is all nice and such, but how do you integrate that with the process of continuously delivering code?

Conventional Commits

Conventional Commits is a convention for how to turn human written commit messages into machine readable meaning. It also happen to nicely follow the convention for SemVer.

  • Major / Breaking Change
refactor!: drop support for Node 6

BREAKING CHANGE: refactor to use JavaScript features not available in Node 6.
  • Minor / Feature
feat(lang): add polish language
  • Patch
fix: correct minor typos in code

GitVersion

GitVersion looks simple and promising at first glance.

However, when digging into the documentation, a different picture immerges.

GitVersion comes with three different strategies; Mainline Development, Continuous Delivery and Continuous Deployment, all quite difficult to decipher and grasp. All of them seems at first to be a possible fit for Trunk Based Development, but I ended up only getting Continuous Delivery to follow the simple philipsopy of Trunk Based Development, yet being flexible enough of when to actually bump versions and not bump any of major/minor/patch more then once until merge to main happens.

Configuration

GitVersion uses a config file to configure strategies, and there are quite few configurable aspects of said strategies. Let’s get started!

Branches

Trunk Based Development is all about getting into trunk (main/master in git lingo) fast. We have two branch strategies; either we are on main (where we don’t do any development) or we are on a development branch. GitVersion however comes with a lot of them that needs to be disabled.

  feature:
    # Match nothing
    regex: ^\b$
  develop:
    # Match nothing
    regex: ^\b$
  release:
    # Match nothing
    regex: ^\b$
  pull-request:
    # Match nothing
    regex: ^\b$
  hotfix:
    # Match nothing
    regex: ^\b$
  support:
    # Match nothing
    regex: ^\b$y

We consider all branches except main as development branches. These should incremented with patch by default.

  development:
    increment: Patch
    # Everything except main and master
    regex: ^(?!(main|master)$)

Detecting Conventional Commits

The following regular expressions detect when to bump major/minor/patch, and are based on the Conventional Commit v1.0.0 specification.

  • Major
    major-version-bump-message: "(build|chore|ci|docs|feat|fix|perf|refactor|revert|style|test)(\\([a-z]+\\))?(!: .+|: (.+\\n\\n)+BREAKING CHANGE: .+)"
  • Minor
    minor-version-bump-message: "(feat)(\\([a-z]+\\))?: .+"
  • Patch
    patch-version-bump-message: "(build|chore|ci|docs|fix|perf|refactor|revert|style|test)(\\([a-z]+\\))?: .+"

Here’s the complete config:

mode: ContinuousDelivery
# Conventional Commits https://www.conventionalcommits.org/en/v1.0.0/
# https://regex101.com/r/Ms7Vx6/2
major-version-bump-message: "(build|chore|ci|docs|feat|fix|perf|refactor|revert|style|test)(\\([a-z]+\\))?(!: .+|: (.+\\n\\n)+BREAKING CHANGE: .+)"
# https://regex101.com/r/Oqhi2m/1
minor-version-bump-message: "(feat)(\\([a-z]+\\))?: .+"
# https://regex101.com/r/f5C4fP/1
patch-version-bump-message: "(build|chore|ci|docs|fix|perf|refactor|revert|style|test)(\\([a-z]+\\))?: .+"
# Match nothing
no-bump-message: ^\b$
continuous-delivery-fallback-tag: ''
branches:
  development:
    increment: Patch
    # Everything except main and master
    regex: ^(?!(main|master)$)
    track-merge-target: true
    source-branches: []
  feature:
    # Match nothing
    regex: ^\b$
  develop:
    # Match nothing
    regex: ^\b$
  main:
    source-branches: []
  release:
    # Match nothing
    regex: ^\b$
  pull-request:
    # Match nothing
    regex: ^\b$
  hotfix:
    # Match nothing
    regex: ^\b$
  support:
    # Match nothing
    regex: ^\b$

Tagging

A common practice when releasing in git is to tag a commit with the release number, i.e. v1.2.3. When practicing continuous delivery, this can become quite verbose, especially when producing pre-releases, as every commit becomes a potential release. GitVersion Continuous Delivery strategy traverses each commit from last version tag when calculating the current version, so pre-releases does not necessary need to be tagged. Instead, the commit sha can be used as the pre-release identifier for back-referencing any pre-release artifacts.

Ofcourse, when using Github or similar platforms, a release is a tag, so in that case you might want to use tagging for pre-releases anyway.

Caveats

Forgetting to merge a merge commit back to the development branch when continuing working on the same branch means the automatic patch increment that happens for commits after branching does not occur. This will skew any pre-release versions. According to the documentation, the parameter track-merge-target should solve that scenario, but it seems it has not been implemented!

Summary

GitVersion turned out to be a really nice versioning tool once overcoming the knowledge curve of understanding all the concepts, features and configurations. Check out the GitVersion Github Action and Azure Pipeline Task for simple integration!

Leave a Comment