2007-11-14 QA Meeting Recap

John Poelstra poelstra at redhat.com
Thu Nov 15 03:54:47 UTC 2007


shrek-m at gmx.de wrote:
> John Poelstra schrieb:
> 
> ---- snip ----
> 5) Bugs previously opened against test releases will also be merged into
> the released version. For example:
>            \
>    FC6Test1 \
>    FC6Test2  | ===> 6
>    FC6Test3 /
>    FC6     /
> 
> 
> 
> think about what kind of bugs and what amount you will find under the
> final/stable release version which was rawhide and not the final release.
> 
> sorry but
> fc6test1  (+ devel updates)
> fc6test2  (+ devel updates)
> fc6test3  (+ devel updates)
> == rawhide and not 6
> 

I agree there isn't 100% correct.  However, that is 2 releases ago... 
those bugs are more likely to be associated with FC6 then with today's 
rawhide.  I believe calling them 6 makes more sense.  It also makes them 
easier to track down as likely candidates for flipping to 
"NEEDINFO--please test against current release" and auto-closing if 
there is no response [1].

> 
> f9alpha  (+ devel updates)
> f9beta  (+ devel updates)
> f9preview  (+ devel updates)
> == rawhide and not 9

Agreed

> f9_the_last_releasecandidate  ==  9

Agreed

[1] All of the above raises the question of what kind of processes we 
should have around managing and triaging open bugs.  I don't believe 
what is proposed above affects the future so much as cleans up the past.

Should all the open bugs that happen to have a version of "rawhide" at 
the GA of Fedora 9 stay "rawhide" forever, or should they be mass-moved 
to "9"?

What do people think?

John




More information about the test mailing list