Releases/Thunderbird3.0a1/Post Mortem: Difference between revisions

no edit summary
No edit summary
Line 4: Line 4:


* QA was quick!
* QA was quick!
** Took 3 days, finding [[QA/Thunderbird3/TestResults/Alpha1|~20 bugs]].
*** No showstoppers.
** A substantial amount of stress on the QA folk(s), though nothing overwhelming.
** Bugzilla flags have to be disabled post-a1-release - {{bug|434956}}


* No RC2 needed
* No RC2 needed.
** Alpha 1 was stable for an alpha.
** Whatever unstable implementations of code had already been backed out.
** Aug 2005 to end-April 2008 code -> 32 months since TB 2 codebase branched.


== What could have gone better, and lessons learned ==
== What could have gone better, and lessons learned ==


* Learned too late about branding subtleties.
* Learned too late about branding subtleties.
(List bugs here)
** {{bug|430762}}
** {{bug|432188}}
** {{bug|433630}}
** {{bug|432756}}?
** ...


* Notified IT too late about need for MoMo svn up on production site
* Notified IT too late about need for MoMo svn up on production site
** {{bug|433593}}
* Not much community involvement for alpha 1 testday.
** Have to reach out more to the community besides existing efforts.
** nth10sd - My hands were tied on this; working on the testplan & almost nobody showing up to help QA testing.
*** Had been in over-drive to finish up QA alone in this regard, but not overwhelmingly stressful though.
** We really should involve more of the community, even regulars popping by in the channel discussing certain areas of testing collaboratively would be helpful.
** Alpha 2 QA plan?
* No significant numbers of Litmus testcases for Thunderbird 3 testday
** Litmus admin was on leave, so the TB 2 testcases couldn't be copied in time.
*** TB2 testcases were used instead.
** Now in progress - {{bug|431883}}
canmove, Confirmed users
1,008

edits