lubuntu-qa team mailing list archive

[C de-Avillez <hggdh2@xxxxxxxxxx>]

On Tue, 4 Sep 2012 17:16:57 -0500
C de-Avillez <hggdh2@xxxxxxxxxx> wrote:

> On Tue, 4 Sep 2012 18:35:58 +0100
> Phill Whiteside <PhillW@xxxxxxxxxx> wrote:
> 
> > Hi folks,
> > 
> > recently a bug reported by a QA member was marked as a duplicate (no
> > harm there), but the 'master' bug was set as 'private'. This meant
> > that the person who registered it could not access it to view /
> > update etc.
> > 
> > I had a chat with the bug squad about this. If this happens they ask
> > that you go to #ubuntu-bugs and raise it with them. The bot does go
> > trawling for duplicated bugs, but it is not aware if the bug it is
> > using as master is private. I'll get the wiki area updated.
> 
> To expand a bit more on it: there are two major types of private bugs:
> 
> (this is centered on Ubuntu, but should be similar on other projects,
> like Lubuntu)
> 
>  * born private (usually security and apport crashdump bugs)
>  * made private later on.
> 
> The apport crashdumps are born private because they carry a coredump
> file -- and it can carry a LOT of private data in it. Additionally,
> apport processing runs a GDB 'bt full' for both the offending thread
> (the one that got hit) and for all threads. The 'bt full' sports
> variable assignment values. Depending on the program, and where it got
> hit, these variables values could also hold private data. No matter
> what, at the end of apport processing, the coredump is deleted.
> 
> So... apport will, usually, not make a crashdump bug public. But it
> knows about it, and will match as needed, marking newer bugs as
> duplicates.
> 
> A subset of Launchpad users have access to these bugs, and can
> manually review and decide on an action -- the worst scenario is we
> see private data, and will have to discuss with someone else (a
> maintainer/developer of the package) if the private data is critical;
> if it is not, we can edit/remove the offending pieces, and make the
> bug public. If it is, on the other hand, there is not much we can do.
> But these are very unusual cases, I do not remember seeing one myself.
> 
> So, to summarise: yes, your bug can be dupped to a private bug. In
> this case, please pop in #ubuntu-bugs, or email Bug
> Control/Ubuntu-QA, and ask for someone to perform the check. We will
> be happy to help.
> 
> It is just that there are more bugs than the time we have...

Darn, forgot the most important piece: Thank you, Phil, for pinging us
and taking the lead on making this public :-)

Attachment: signature.asc
Description: PGP signature