Updates Criteria Summary/Brainstorming

Till Maas opensource at till.name
Mon Nov 22 18:19:05 UTC 2010


On Mon, Nov 22, 2010 at 09:15:17AM -0800, Adam Williamson wrote:
> On Mon, 2010-11-22 at 18:09 +0100, Emmanuel Seyman wrote:

> > The whole 'push directly to stable' arguement rests heavily on the principle
> > that an update is always better (from a QA standpoint) than whatever it's
> > replacing. The problem is that there's no way to guarantee this, essentielly
> > because it isn't true.
> 
> I believe Kevin would say his position is that the update is better than
> what's there already *sufficiently often* that allowing unrestricted
> updates is a net benefit (the question is whether an occasional bad
> update is a worse problem than some updates being delayed for a week or
> longer in the case of untested critpath updates).

It is not some update that is delayed, but one that fixes a very bad bug
like e.g. a remote code execution vulnerability. And the worst update is
afaik that people had to use the command line to update instead of being
able to use packagekit or kpackagekit.

Regards
Till
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: not available
Url : http://lists.fedoraproject.org/pipermail/devel/attachments/20101122/7ee54af9/attachment.bin 


More information about the devel mailing list