User:Greenfeld/Testing Information

Testing Information

This page is meant to help those who are new to testing the XO laptop. It provides links to a variety of resources which may be useful.

XO Hardware Versions

There are basic three versions of the XO hardware currently accessible to testers. They are the XO-1, XO-1.5, and XO-1.75.

DESCRIBE THE THREE XO TYPES HERE.

Each of these basic versions may very in terms of:

• Keyboard controller: XO-1s may have ALPS or EnE keyboard controllers. XO-1.5 & 1.75 99% of the time will have EnE keyboard controllers, unless they are placed into an XO frame with an ALPS keyboard base. It is unlikely you will encounter an ALPS keyboard unless you have a very early XO-1.
• Keyboard language: Multiple keyboard languages are supported. For a complete list look at Keyboard#Languages_other_than_English.
• Keyboard style: Both membrane based as well as mechanical ("High School") keyboards are available.
• Touchpad style: ALPS keyboards have a three-segment touchpad, in which only the middle segment senses fingers, and the rest can detect stylus input when enabled. All newer systems only have a single-segment touchpad which senses fingers.
• Touchpad type: Since the ALPS style was discontinued from early XO-1's, Synaptic touchpads well as AVC/Sentelic touchpads have been used.
• Memory/Storage: Deployments may request different amounts of RAM or onboard storage from their default values, and development machines may vary as well.

Bug/Issue Tracking

OLPC vs. Sugar Labs vs. Fedora

There are three bug tracking systems related to OLPC software releases.

• OLPC has its own bugtracker at http://dev.laptop.org/, which also hosts the Trac software repository for OLPC specific software changes. If a bug is found in the custom Linux kernel, a video driver, or any olpc-specific package, the problem should be filed in this system.
• The main user UI for children is called "Sugar". Sugar Labs is the group responsible for maintaining this. If a bug is found in the Sugar UI, or any Sugar-startable Activity, it should be filed at http://bugs.sugarlabs.org/.
• At its core, XO software releases are based on Red Hat's Fedora Linux distribution, and OLPC software releases are officially considered a "Fedora remix". But the Linux kernel, Sugar packages, and Sugar activities included with OLPC software builds typically differ from what Fedora provides by default.
  Bugs in Fedora are logged at https://bugzilla.redhat.com/, but please be sure that you are not using a custom OLPC package or installation approach which is causing the problem before filing the issue upstream.

When in doubt, please file a bug for any problem found using an XO on the OLPC bugtracker, and someone will provide information on where it should go as appropriate.

Where should I file bug for...

• The Base Sugar UI: http://bugs.sugarlabs.org/
• Sugar Activities (Browse, Maze Write, etc.): http://bugs.sugarlabs.org/
• GNOME Programs (Abiword, Audacity, etc.): https://bugzilla.redhat.com/
• Power management issues:
• Linux kernel messages:

Test Cases

Test Case System

The Test cases page has a complete list of test cases stored in this wiki. Many of these are out of date, and some are no longer relevant but for the most part the differences between releases can be easily figured out.

This is part of a test case management system which was developed within the wiki. However it tends not to be used for logging results nowadays because:

• It is hard to handle test case scheduling with this system.
• There is no (known) easy way to see what has not been run yet against a build.
• MediaWiki caches wiki pages, even those which use semantic data like the test case system. This means every wiki page in the TCMS shows stale results until the cached copy expires, or a page purge is forced.

Smoke Tests

The latest version of the "one hour" smoke test can be found here.

Collaboration Testing

A few of the test cases as well as the smoke tests require more than one XO to test collaboration between laptops. If you do not have more than one XO to work with, then you can safely skip these sections (but log you did if reporting results).