From Fedora Project Wiki

< User:Roshi(Redirected from Test Cartography)

This is a place-holder page for "Testing Maps" documentation. More to come. This is all very unstructured right now. The goal is to make our testing efforts similar to a Gamebook - instead of making decisions for the protagonist, what hardware you have and what tests you've done determine the next testcase to run.

Example

Below is a rough example of how I decide which tests to do as I work through the matrices.

 Arch? x86_64
   EFI Install - https://fedoraproject.org/wiki/QA:Testcase_USB_stick_DVD_dd
 Installation Type?
   Graphical - https://fedoraproject.org/wiki/QA:Testcase_Anaconda_User_Interface_Graphical
   Text - https://fedoraproject.org/wiki/QA:Testcase_Anaconda_User_Interface_Text
   VNC - https://fedoraproject.org/wiki/QA:Testcase_Anaconda_User_Interface_VNC
 What kind of drive?
   SATA Drive - https://fedoraproject.org/wiki/QA:Testcase_install_to_SATA_device
 What kind of partitioning?
   Autopart - https://fedoraproject.org/wiki/QA:Testcase_Anaconda_autopart_install
     Data on the Drive?
       Yes - Reclaim Space - https://fedoraproject.org/wiki/QA:Testcase_Anaconda_autopart_(reclaim)_install
       No - Free Space - https://fedoraproject.org/wiki/QA:Testcase_Anaconda_autopart_(use_free_space)_install
 Encrypting partitions?
   Base Setup - https://fedoraproject.org/wiki/QA:Testcase_base_startup
 Creating a user during Install?
     Yes - Continue testing...
     No - Initial Setup - https://fedoraproject.org/wiki/QA:Testcase_base_initial_setup
 Desktop Environment
     GNOME or KDE
       Login - https://fedoraproject.org/wiki/QA:Testcase_desktop_login
       Panel Basic - https://fedoraproject.org/wiki/QA:Testcase_desktop_panel_basic
       Etc...

Testcase Meta Data

One issue that arises when looking at grouping test cases like this, is how do you add test cases? If this is a manual process, then the work to add a test case could large (depending on how many test maps you had). Manually doing things is hardly what I'm arguing for. I think that by adding some metadata to our test cases we can group these tests relatively easily.