Broken deps in the stable release are not acceptable

[Christopher Aillon]

On 12/29/2007 04:17 PM, Rahul Sundaram wrote:
> Christopher Aillon wrote:
>> On 12/29/2007 02:32 AM, Rahul Sundaram wrote:
>>> Christoph Wickert wrote:
>>>> I completely agree with you. Maybe we could say that updates are 
>>>> allowed
>>>> to bypass testing if they fix
>>>> a) serious bugs
>>>> b) bugs marked as "urgent"
>>>> c) broken deps
>>>
>>> b) isn't a good criteria since anybody can mark any bug as urgent. If 
>>> the priority field in bugzilla is restricted to package maintainers 
>>> and triagers, I would agree with you.
>>
>> The same maintainer who marks "push right to stable" can tweak the 
>> field before they submit the update and you won't have solved anything.
> 
> Even if it had a strict set of rules and maintainers are going to abuse 
> the system,

Hey dude, I wasn't the one agreeing with a set of rules, that was you. 
I'm just saying it's unwise to agree with a set of rules that can still 
be worked around easily.

> they can mark any update as a critical security update and 
> push it through too but then it is much more easier to point out who is 
> responsible compared to users just marking a random bug as a high 
> priority one.

I just noticed that nobody sent out a FESCo Meeting Summary for 
2007-09-27[1].  There, we approved 
http://fedoraproject.org/wiki/LubomirKundrak/SecurityUpdateProcessDraft 
so the Fedora Security Response team would have to approve it before it 
gets released as a security advisory.

[1] At least there's a log at 
http://bpepple.fedorapeople.org/fesco/FESCo-2007-09-27.html

Nobody's implemented that yet, though... Luke?  This would be quite nice 
to get done... :-)