REGRESSION [hardy-backports]: flashplugin-nonfree 10.0.1.218+10.0.0.525ubuntu1~hardy1 causes Firefox to crash on numerous popular websites

[Scott Kitterman]

On Sat, 12 Jul 2008 16:38:38 -0300 "Cody A.W. Somerville" 
<cody-somerville at ubuntu.com> wrote:
>On Sat, Jul 12, 2008 at 4:30 PM, Scott Kitterman <ubuntu at kitterman.com>
>wrote:
>
>> On Sat, 12 Jul 2008 16:11:32 -0300 "Cody A.W. Somerville"
>> <cody-somerville at ubuntu.com> wrote:
>>
>> <snip>
>
>> >Anyhow, I suppose the first step is blocking the update and getting a 
fix
>> out. After, we can look at how this happened and what we need to do to
>> reduce the risk of something similiar occuring again.
>> >
>> ....
>> Everyone knows backports are crack and so ....
>>
>> I am aware (I approved the backport) and as soon as I get somwhere that
>> lends itslef to a stable IRC presence (less than an hour) I intend to 
hunt
>> down an archive admin and deal with it.
>>
>> Backports by design are risky and aggressive.  Sometimes they don't work
>> out and we deal with it.  We actually don't want to prevent risk taking 
in
>> backports, so please don't 'fix' the process.
>
>
>If not the "process" then probably the documented policy (which seems to be
>https://help.ubuntu.com/community/UbuntuBackports). In the "Stability"
>section it says: "A great deal of work goes into testing backports and it's
>highly unlikely for a backport to be a severe regression from the version 
it
>replaces."
>
>Furthermore, since it seems the common user enables -backports, we should 
do
>a better job of informing them of the risks - probably including a warning
>when actually enabling it.

I've ack'ed about 3 dozen backports requests in that week or so.  This is 
the only one I'm aware of to have a regression.  In fact this is only 
second regression I've ever had from a backport I approved.  I think highly 
unlikely is accurate.

I have some ideas for myself as an approver to prevent a recurrence, but I 
think the general process/docs are fine.

Scott K