Printer-friendly version

Author(s): Ara Pulido

The proprietary software industry makes heavy use of automated testing tools to thoroughly test their software against the original design specifications before release, in part because they often do not have access to the number of voluntary testers relied on by the open source community.

While there is a wide selection of proprietary GUI test automation tools, the selection of Free tools is more limited. The preferred approach in GNOME is to use the accessibility technologies to access the UI information. There are several test frameworks (Dogtail, Strongwind, LDTP) available that use essentially the same approach but to date there has been little progress on forming a central testing effort for GNOME.

The GNOME Desktop Testing project aims to implement a testing framework on top of LDTP that will make it easier to write new test scripts for GNOME that are consistent, reusable and maintainable. The framework has 3 different parts: a desktop testing library, a test runner and the tests themselves.

• Desktop testing library: a python library that creates a wrapping framework for GNOME applications based on LDTP that will hide the lower levels of details to the test script.
• The Test Runner parses the test suites description (and XML file), takes care of running the test cases and creates a XML log and a HTML log (human readable).
• Tests: The scripts themselves.

The talk will cover the basics of the GNOME Desktop Testing project (http://live.gnome.org/DesktopTesting) and a 5-10 minute tutorial.

Author bio: Canonical

Ara Pulido is member of the Ubuntu QA Team, working for Canonical. Her main duty in the QA Team is testing automation tasks. She is the lead of the GNOME Desktop Testing project (http://live.gnome.org/DesktopTesting).