action #1310
closedImprove debugging capabilities for the tests
100%
Description
Once the test is running there is no way to know what is happening inside the test, or have a good knowledge of why the test fails. We need to provide tools to debug the test. These tools can include a debug console to print logs, or a way to inspect the execution of the test (traces, break points, etc).
Another solution is to integrate into the final video a WebVTT subtitle overlay to show the keystrokes in real time, allowing people to better understand what happens, when and why.
And a way to run again the tests, using a previous VM state
Updated by RBrownSUSE about 9 years ago
- Category changed from 138 to Feature requests
Updated by okurz almost 9 years ago
- Status changed from New to In Progress
aplanas wrote:
Once the test is running there is no way to know what is happening inside the test, or have a good knowledge of why the test fails. We need to provide tools to debug the test. These tools can include a debug console to print logs, or a way to inspect the execution of the test (traces, break points, etc).
Nowadays the logs include a lot of information from the executed tests including asserted screens and key presses.
And a way to run again the tests, using a previous VM state
Possible and properly documented, e.g. see the 'SKIPTO' flag and 'MAKETESTSNAPSHOTS' for creating qemu snapshots and restoring from them for fast re-execution. I consider this done.