QA/Execution/Meetings/2009-11-18/MeasuringQuality: Difference between revisions

Jump to navigation Jump to search
 
Line 78: Line 78:
* Increase automated coverage for core features, to free up time to test manually (ad-hoc)
* Increase automated coverage for core features, to free up time to test manually (ad-hoc)
* Ask development and product managers to look/talk over your current coverage
* Ask development and product managers to look/talk over your current coverage
==Henrik==
* We need a central place people can use to send and view reports. That will allow us to start a collaboration between all active users
* Having an ongoing review cycle for tests to land in the repository
* Tests should absolutely be runnable with current Mozmill release. There should not be a need to clone the Mozmill repository. Users can get annoyed.
* Reducing the amount of external content to lower the number of
failures due to website changes. With that we can work on real failures and bugs.
* Running tests on a daily basis (smoketests/bft/update) to add more
trust and catching regressions earlier
* In-detail project documentation (overview, roadmap, API's, ...) and details of ongoing projects
* Only run Mozmill tests which are not broken due to any known bug of Firefox. It will reduce the noise of broken tests.
* Making current status more prominent and talk about it in regularly meetings
(testdev, dev meeting?)
* Get attention of developers for bugs found with Mozmill
* Getting more contributors on different platforms
* Actively use the Mozmill list to keep people up to date


== Discussion Notes ==
== Discussion Notes ==
Line 103: Line 119:
** QA to get involved back into triage meetings, actually look at those bugs.  start asking questions on those bugs.  push for those on 3.7 community triage meetings.  And for mobile also.
** QA to get involved back into triage meetings, actually look at those bugs.  start asking questions on those bugs.  push for those on 3.7 community triage meetings.  And for mobile also.
** We dont have a good grasp on how many testcases are going on, how many testcases are covered there.  Having a way to measure how many devs have unit tests, how many we have functional tests...  can we work toward that?
** We dont have a good grasp on how many testcases are going on, how many testcases are covered there.  Having a way to measure how many devs have unit tests, how many we have functional tests...  can we work toward that?


= Draft 2 (Summarized) =
= Draft 2 (Summarized) =
canmove, Confirmed users, Bureaucrats and Sysops emeriti
4,747

edits

Navigation menu