Configure Developers and Teams

Your knowledge maps are based on colors to give you an accessible high-level overview. The system will automatically assign a distinct color to each top contributor in your codebase on the first analysis. But you still need to assign a color to the other developers if you want to identify them in the visualizations. Similarly, you can define teams of developers and assign each team a color as well.

Sample on colored knowledge maps

Fig. 181 Sample on colored knowledge maps.

The rest of this guide will walk you through the configuration.

Important: Run an Initial Analysis Before You Configure Authors

CodeScene mines a list of all contributing developers. Note that this list is mined and updated during each analysis. That means you need to run one initial analysis before the tool gives you the option to configure developer properties!

Define Your Development Teams

Click the Teams tab in your project configuration to proceed to the teams configuration, as shown in Fig. 182.

Configure teams for a project in the Teams tab.

Fig. 182 Configure teams for a project in the Teams tab.

For each team in your organization, specify the following properties:

  • Team name: This will be used to identify the team. Later, when you configure developers, you’ll assign them to the team names you chose here.

  • Team Color: The team color is used in the visualization of knowledge distribution on team level.

Add a team for each team in the organization that works on your codebase.

Configure the teams that reflect your organization.

Tip: Some organizations just use one development team. In that case, introduce virtual teams that depend upon the responsibilities of the different developers. For example, you might want to define a Feature team, a Maintenance team and an Infrastructure team. Using this strategy, you’d be able to identify code at risk for incompatible parallel changes since different forces lead to the changes.

Even Open-Source Software Has Teams

The team definition is straightforward if you analyze a codebase that’s owned by a traditional organization; Just use the information from your organizational chart. However, we find it interesting to apply teams to open-source codebases as well.

So if you happen to analyze an open-source project, consider introducing the following teams to get additional social information:

  • Define a teams for the organization that owns the code. For example, if you analyze the Clojure codebase, you’d define Cognitect as one team. If you analyze one of Microsoft’s open-source codebases, you’d use Microsoft as one team.

  • Define a team for third party developers that contribute to the codebase

  • Consider defining a team of the core maintainers too.

Configure Developer Properties

The developer properties are a bit more tricky than the team configuration, so please let us walk you through them one by one as illustrated in Fig. 183.

Clone developer information

Fig. 183 Specify organizational information for each developer.

CodeScene automatically updates the list of contributing developers; If a new developer starts to contribute code, they’ll be present in the list and the tool lets you configure their properties.

Here are the properties you need to specify:

  1. Color: The color is used on the knowledge maps to uniquely identify a developer. Try to assign the top contributors as distinct colors as possible.

  2. Active/Ex-Developer: By default, all developers are considered active. If some of them leaves your project, mark them as Ex-Developers and CodeScene will include them in the Knowledge Loss Analysis.

  3. Team: The third column lets you assign the developer to a team. That will include them in the Team Knowledge Distribution Analysis.

  4. Exclude author from analysis: If you check this option, the author will be excluded from all social analyses (although their contributions will still be included in the technical analyses like Hotspots and Code Churn). This is an option you use in case you have roles like System Integrators that only merge code, but never actually make their own contributions.

Once you’ve defined all developer properties you just need to run a new analysis and you’ll get a smorgasbord of interesting social analysis results.

Import a Definition of Development Teams

It may well be impractical to configure each team and developer via the UI, particular for large organizations. That’s why CodeScene supports import of the organizational chart.

You find the import functionality in the Team configuration:

Import developer information

Fig. 184 Importing developer information by uploading a CSV file.

The input file specifies your organization. The file format has to be a CSV with two columns: author and team.

Authors and their Aliases: Mapping Version-Control Names to People

Often, over the lifetime of a project, some developers will sign their commits with different names. This can be a source of inaccuracies for CodeScene’s social analysis tools.

To deal with this, CodeScene provides an interface that allows you to specify the version-control names that correspond with real people. In CodeScene, when we talk about an Author, we mean the real person. Team membership, author exclusion and ex-developer status belong to the developer. Each developer has at least one Alias, which is how they are identified in version control.

For example, an author named Jane Doe might have several aliases in version control commits: Doe, Jane, janedoe, J. Doe, etc. This interface allows theses aliases all to refer to the same person, which provides more meaningful results in social analyses and unifies the information we have about the author in question.

Workflow

To access the interface, click on the Developer identity mapping tab in the Teams/Developers configuration corresponding to the project you are interested in.

On the left, the interface displays a list of the current authors.

The developer list

Fig. 185 The Author panel

Choose an author you want to work on. This is the name that you want to keep.

On the right, a new list will appear.

Developer list with alias list

Fig. 186 The alias list.

To find the alias you want to merge, you can use the “Filter aliases” box to search for matching aliases. Regular expressions are allowed, with whitespace counting as a logical OR.

Use the filter box to search for aliases

Fig. 187 Using a regular expression to filter aliases.

When you have found the alias or aliases you are looking for, select them on the right and click on Stage changes.

Developer list with alias list, getting ready to select.

Fig. 188 The alias list, ready to merge.

In this example, we want to merge the aliases “Calvin” with “Calvin Buckley”. This will update the list of authors on theleft. Now we can either make other modifications or click on “Submit” at the top of the window to finalize the operation.

Separating aliases from their authors

If we change our minds, we can later separate these aliases from the author that we assigned them to. To do this, we select the corresponding checkbox and click on Stage changes again.

Selecting an alias to unmerge

Fig. 189 Separating an alias from a developer.

After clicking on Submit, the aliases we chose to separate will become full-fledged authors. Because these are new identities, group membership, ex-developer status, and exclusion status will be lost. Merging and unmerging an alias is lossy.

Renaming authors

Because an Author is separate from the version control alias, authors can be renamed without changing how they are detected by CodeScene’s analyses. To change a name, click on that author in the left column. You will notice an “Edit name” link next to their name in the box on the right.

The rename field

Fig. 190 Editing a developer’s name

When you run a new analysis, the new name will appear in the analysis results.