[Libreoffice] Release criteria
Michael Meeks
michael.meeks at novell.com
Mon Nov 29 03:10:09 PST 2010
Hi there,
On Sat, 2010-11-27 at 06:57 +0100, Jean-Baptiste Faure wrote:
> > So - I tend to think that time will tell if we need to change this. The
> > seven day period seems long to me.
>
> Seven days is a not so long period when you have to organize manual testing.
> As you can see with OOo 3.3.0 the last stoppers have been found by
Sure :-) but what concerns me is the reset of seven days with no bug;
so we find one bug, and we then reset the seven days of testing when we
have included only a trivial fix ? if so, that seems like wasted time to
me. There comes a point when the negatives (of holding back all the
improvements and bug fixes in the new release) outweigh the positives of
waiting until it is perfect :-) Having a hard rule of no blocker bug for
seven days after the last RC seems to ~ignore the other half of that
balance.
> Perhaps, if micro releases are really only bug fix.
> But that need to be clearly explained to the users and sys-admin in
> charge of deployment of LibO.
True.
> It is very important that the user installing the next release of LibO
> can say: Oh there is less bug in LibO than in ...
Heh - and of course we have to be careful about that.
Thanks !
Michael.
--
michael.meeks at novell.com <><, Pseudo Engineer, itinerant idiot
More information about the LibreOffice
mailing list