Task Repository Connectors

Mylyn allows you to collaborate on tasks via a shared task repository, also known as bug tracking systems. In order to collaborate, you need to have a Connector to your particular repository.

Presently Bugzilla, JIRA, and Trac are supported. To connect to unsupported repositories, see Generic Web Repository Connectors. Also, be sure to vote for your favourite Connector to see it supported earlier, or create a new bug if your issue tracker is not listed.

Bugzilla Connector

Once Mylyn is installed there are a few steps involved to get up and running.

Repository Configuration

See also Bugzilla Connector Troubleshooting.

Bugzilla Tasks

Tips for server administrators

Mylyn periodically checks config.cgi to retrieve the repository configuration. On Eclipse.org this resulted in heavy CPU Load for the regeneration and a big surge in band width use.

Mylyn has been modified to accept gzip encoding on all requests, and will do content negotiation. See bug 205708.

To add caching for your Bugzilla repository:

 via CVS: :pserver:anonymous@dev.eclipse.org:/cvsroot/technology/org.eclipse.phoenix/infra-scripts/bugzilla/
 or: http://dev.eclipse.org/viewcvs/index.cgi/org.eclipse.phoenix/infra-scripts/bugzilla/?root=Technology_Project

Trac Connector

The Trac connector offers two access methods:

The Trac connector integrates Trac queries into the Task List. If you do not know your Trac repository version use the Automatic setting and click Validate Settings.

New Task Editor

A rich editor for creating new Trac tasks is available for repositories that use XML-RPC (see the FAQ for XML-RPC configuration instructions).

Rich Editor, Attachments and Offline support

The Trac connector supports the rich task editor. Attributes and comments can be viewed and edited offline, synchronization is done in the background, and attachments can be posted and retrieved.

Task Context attachments are supported via the context menu actions in the Task List. This support requires the Trac XML-RPC plug-in to be enabled and the integration will fall back to the web mode if it is not, (see: Trac Connector troubleshooting).

JIRA Connector

See this page for install instructions and additional screen shots. Note that the Mylyn JIRA connector requires at least JIRA version 3.3.3 and SOAP/RPC services have to be enabled on the server.

Rich Editor

The JIRA connector provides a rich editor, offline editing, and change notifications.

Search Integration

Searching through JIRA repositories is integrated with the Search dialog.

Query

The JIRA query dialog has been streamlined into a single page. Date range queries are now supported.

See also JIRA Connector Troubleshooting.

Generic Web Repository Connector

The generic web repository connector is NOT part of the default Mylyn install. You can install it from a separate extras update site. See Mylyn download page for more details.

The web connector allow to retrieve tasks from repositories that don't have rich connectors, but can show list of tasks on the web UI. Out of the box connector provides configuration templates for the following issue tracking systems:

Lists of issues can be extracted from existing web pages using simple parsing configuration. Configuration can be also parametrized to make it easier to customize it for a specific project.

The parameters used for configuring project properties are typically substituted into the URLs used to access the repository. Substitution and matching rules can be edited under the Advanced Configuration section on both the Repository Settings page and the Edit Query page.

See FAQ for the troubleshooting tips.

For example, consider the configuration steps for GlassFish project at java.net:

1. Create new Generic web-based repository (in the Task Repository view). GlassFish is using IssueZilla and has a preconfigured template that can be selected by server url https://glassfish.dev.java.net/issues . You can also specify all fields manually in the Advanced Configuration section. For GlassFish the following settings are required:

Note: Query Pattern field should be a regexp with 1st matching group on Issue ID and 2nd on Issue Description. Alternatively, you could use named matching groups: ({Id}.+?), ({Description}.+?), ({Status}.+?), ({Owner}.+?) and ({Type}.+?), then they can appear in query regexp in an arbitrary order. The second option requires build 2.0.0v20070717 or later.
Note: the above fields are using parameter substitution ${..}. Variables serverUrl, userId and password are substituted from the values of corresponding fields of the repository preference page. In addition you can specify any arbitrary parameters and their values that will be also substituted into the template fields.
Note: the SourceForge template included with connector assume that single repository is used for all projects. User should create multiple queries, and set project parameters at the query level. Because web connector don't support actions like "open repository task" there is really no need to create separate repositories per project and if you think about it that is how it work for connectors for Bugzilla and JIRA. However, it is still possible to setup separate repository per project using repository url like http://sourceforge.net/tracker/?group_id=172199 and accordingly updating derived urls is the advanced repository settings.

For the web repository that require user to login, use advanced configuration in following way. This configuration is for GForge, you might need to change it for other repositories:

2. Create a new query for the GlassFish task repository created above (either from popup context menu in the Task List view or using a "New..." wizard from File -> New... -> Other... menu).

XPlanner Connector

The XPlanner connector is currently in the Experimental state. It supports XPlanner version .7.

Query

XPlanner Connector supports creating queries from your XPlanner repository. You can select to create a query for your own in progress tasks (the default), or select projects or iterations or user stories from a list. You can also select ones assigned to you, or all. Currently you can choose to group the results by tasks or user stories. If you group the results by tasks, this will result in a single query with the applicable tasks underneath it. If you choose to group the results by user stories, this will result in multiple queries added to the task list, each one corresponding to a user story, with the applicable tasks underneath it:

Task Editor

A rich editor for editing XPlanner tasks is available. It currently doesn't support all of the attributes you can change in XPlanner itself, but allows you to modify the basic ones necessary for XP development -- e.g. summary, description, estimated time. The attributes section of the editor displays an XPlanner-like view of the meaningful attributes, some of which are not editable.

User Story Editor

If you open an XPlanner element by id, using the "Open Repository Task by Key/ID" dialog, and that id is a user story id, you will see the read-only user story editor. This editor is just useful for seeing the major details of a user story -- no items in it can be changed, since there is no really good mapping between a user story and a Mylyn task.