Report bugs and feature requests

This guide provides a set of rules for reporting issues and managing tickets in Datagrok. Everyone, the Datagrok team and customers alike, reports bugs and feature requests at github.com/datagrok-ai/public/issues. If the fix turns out to be in the core platform, the team links a tracking issue to yours, so you don't need to decide where the problem is.

Reporting an issue

A good ticket has a type, title, and description; the team adds an assignee, priority, and target date. Larger pieces of work are split into sub-issues of a parent issue. Here we provide recommendations for each item:

• Issue type. Use Bug for unexpected behavior, Feature for a new capability or an improvement of existing functionality or UI, and Task for other work.
• Title. Start the title with the platform module or package name, then name unexpected behavior for a bug and a new feature name or improvement summary for an enhancement.

Issue type	Formula	Example
Bug	[Platform module] or [Package name] : [Unexpected behavior]	Line chart: Y-axis label is broken
Feature	[Platform module] or [Package name] : [New feature name] or [Improvement summary]	Bar Chart: context menu harmonization

• Description is important for testing and writing release notes. For an enhancement, write the description of the new functionality. A bug description should include the following items:
  • Instance and its version. If the bug concerns a package functionality, mention the package version.
  • Data on which the bug is reproduced (optional). If a bug occurs on a particular dataset only, include it in the bug description.
  • Steps to reproduce. A step-by-step description of your actions causing the bug (where to click, what to enter into inputs, which checkboxes to mark, etc.)
  • Expected Result. Describe how the platform is supposed to work—the ideal result you want but don't get.
  • Actual result. The result that you receive when using the platform, which you don't expect to receive.
  • Code examples (optional). If the bug concerns working with the platform API, scripts, queries, or something where you use a code, give an example.
  • Attachment (optional). You are welcome to attach screenshots, GIFs, recorded videos, etc. to complete the issue description.
• Priority. The team sets the Priority field. Use your common sense above all rules described below.
  • Urgent. It is a bug in the main functionality of the platform. Such as Chem Descriptors, Detectors, and so on.
  • High. An important request from the customer, something that blocks others' work, or a task with the exact due date. Set the target date for such tasks even if it was not initially specified by the issue's reporter. In most cases, such requests should be resolved in a week.
  • Medium. The default priority for the task.
  • Low. A good enhancement for the platform which does not have a big impact on the users.
• Target date. Set the Target date field according to the priority, client's request, or corresponding workstream date. Generally, medium-priority issues should be done within two weeks, high-priority within a week, and urgent ones within a few days. The release an issue ships in is its milestone.
• Assignee. To understand the person responsible for a particular package, look inside the package.json file, which must exist for each package. Inside this file, there is an "author" field from which you can find out whom to assign.
• Linking (optional). Use Relationships in the issue sidebar to mark an issue as Related to another, in this or any other datagrok-ai repository.

Managing tickets

To track the actual state of the platform, it's important to keep tickets up to date:

• Assign the issue to yourself when you start working on it, and comment on progress or blockers.
• Link the pull request to the issue with Closes #123 in its description, so the issue closes when the pull request is merged.
• Close an issue that won't be done as not planned, or as duplicate with a link to the other issue.

Note: Write meaningful commits, and don't forget to mention the issue number. For details, see the commit message policy.