[Fedora QA] #261: Revise upgrade test case set

[Fedora QA]

#261: Revise upgrade test case set
--------------------------+-------------------------------------------------
  Reporter:  adamwill     |       Owner:  hongqing 
      Type:  task         |      Status:  assigned 
  Priority:  major        |   Milestone:  Fedora 17
 Component:  Test Review  |     Version:           
Resolution:               |    Keywords:           
--------------------------+-------------------------------------------------
Comment (by adamwill):

 The current situation is how it's meant to be - though obviously, since
 it's possible to interpret it as a problem like you did, we need to
 clarify things.

 It's actually okay to have tests in the matrix which are associated with
 no release phase. I call these 'advisory tests'. The idea is that there
 are things we want to test so we catch and file any bugs in them and so we
 know if they're broken or buggy and can document the problems, but bugs in
 them are _not_ release blockers. That's fine.

 That's what the 'upgrade two releases' test is: it's not a release
 blocking issue if it fails, but we do want to know whether there are major
 problems we should document and at least _try_ to fix.

 Obviously, we need to make this more clear somehow, so people don't have
 the same questions about it that you did!

 The current 'official' support is exactly what's in the criteria: upgrade
 from one release ago is release-blocking, upgrade from more than one
 release ago is not.

-- 
Ticket URL: <https://fedorahosted.org/fedora-qa/ticket/261#comment:4>
Fedora QA <http://fedorahosted.org/fedora-qa>
Fedora Quality Assurance