James Laska said the following on 02/09/2009 05:45 AM Pacific Time:
On Mon, 2009-02-09 at 01:17 +0530, Rahul Sundaram wrote:
http://fedoraproject.org/wiki/QA/ReleaseCriteria
I'm of the mindset that QA should not solely define the release
criteria. Defining what a release should be must be a collaborative
decision. Perhaps the above link is something that each team provides
*minimal* content for and FESCO reviews/approves? QA can then taylor
testing around what expectations the development teams are setting along
with the approved release criteria. Should all this data eventually
land in each test plan as 'exit criteria'?
I'd love to see a world where each test team defines their exit criteria
for testing. I find it difficult to fill a single document with
absolutes that properly apply to all levels of testing throughout the
operating system. Success for the Fingerprint feature will look much
different than success for KDE4.2 and kernel.
That doesn't mean that QA cannot or should not be involved in deciding
whether something blocks the release. I wonder if the release blocking
decision could be more collaborative?
I think this last part would be valuable. I'd also like to see more
discussion and notification from QA when a test or final release is
deemed "ready to go"... at a minimum an email to fedora-test-list
stating what has happened or been decided and who has made that
decision. I'm not advocating more "process", but more communication
than an easily missed IRC conversation.
Nothing is worse than spending several hours triaging bugs to a blocker
list only to find out that the blocker list isn't being used or that the
release has been deemed "done enough" and has gone to the mirrors. I
know every bit helps, but it is still demoralizing.
John
--
fedora-test-list mailing list
fedora-test-list@xxxxxxxxxx
To unsubscribe:
https://www.redhat.com/mailman/listinfo/fedora-test-list
