[Fedora QA] #47: Obsolete install test cases for Fedora 13

[Fedora QA]

#47: Obsolete install test cases for Fedora 13
--------------------------+-------------------------------------------------
  Reporter:  jlaska       |       Owner:           
      Type:  task         |      Status:  new      
  Priority:  minor        |   Milestone:  Fedora 13
 Component:  Test Review  |     Version:           
Resolution:               |    Keywords:           
--------------------------+-------------------------------------------------
Comment (by rhe):

 Replying to [comment:3 jlaska]:
 > Replying to [comment:2 rhe]:
 > > Agree.
 > >
 > > Do you think if
 [https://fedoraproject.org/wiki/QA/TestCases/BootMethodsCdrom
 QA/TestCases/BootMethodsCdrom] is already covered by
 [https://fedoraproject.org/wiki/QA/TestCases/InstallSourceCdrom
 QA/TestCases/InstallSourceCdrom], and
 [https://fedoraproject.org/wiki/QA/TestCases/BootMethodsDvd
 QA/TestCases/BootMethodsDvd] covered by
 [https://fedoraproject.org/wiki/QA/TestCases/InstallSourceDvd
 QA/TestCases/InstallSourceDvd]?
 >
 > Good question.  I'm not fully convinced these tests could be
 consolidated since I they do target different areas of the installer.  For
 example, the Boot* tests are designed to validate that the media is
 properly constructed, can boot into stage#1 loader.  The InstallSource*
 tests focus more on whether the installer is able to load the stage#2
 install.img from the specified location.  They are similar in that if you
 boot and install from a DVD, you've tested both scenarios.  But it's also
 possible to boot from {CD,DVD,PXE,USB,grub} and install from
 {network(http,nfs,ftp,nfsiso),CD,DVD,hd,hdiso}.
 >
 > The general test strategy for installation test outlines testing the
 inputs/outputs at each phase of the installer (see
 https://fedoraproject.org/wiki/QA:Fedora_13_Install_Test_Plan#Test_Strategy).
 It's definitely a bit odd at first, but given the complexity of the
 installer, it's proven to be a useful approach.  That said, I'm 100% open
 to devising a new test strategy if there are ideas on improving the
 workflow/procedure.  I'd be hesitant on removing tests that cover
 different areas until there is a larger workflow/procedure defined to
 account for the change.

 I see, the Test Strategy does make sense though a bit odd. :)  Thanks.

-- 
Ticket URL: <https://fedorahosted.org/fedora-qa/ticket/47#comment:4>
Fedora QA <http://fedorahosted.org/fedora-qa>
Fedora Quality Assurance