Installation fails and the corresponding bug report

[Colin Watson]

On Sat, Feb 21, 2009 at 10:15:37AM +0100, Vincenzo Ciancia wrote:
> Colin Watson ha scritto:
> > In future please file a new bug rather than doing that. It makes our
> > lives much more difficult when people tag along to an existing ubiquity
> > bug; it's actually quite difficult for users to tell whether they're
> > experiencing the same cause of problem as somebody else.
> 
> I don't think I am getting the point: usually when I find my bug already 
> reported I add information to it. In this case the bug I found is my bug 
> so why should I open a new bug? Is this just a policy for ubiquity?

The problem is that most people (through no fault of their own; it's
just presumably not something they need to do very often) are *very* bad
at telling the difference between two installer bugs with perhaps
similar symptoms but different causes. I frequently have problems where
people decide that every partitioning bug is the same ... Perhaps the
installer suffers from this particularly because it's harder to go back
and try different things, or perhaps just because Ubiquity's tracebacks
have historically not been as distinctive as we might like - at the
moment it's entirely possible for completely different bugs to show up
with the same traceback, and it's usually necessary to look at the full
logs to distinguish them.

It is much more effort to split a bug apart once ten people have
dogpiled on it with a mix of symptoms than it is to mark ten bugs as
duplicates. In the former case, there's usually nothing for it but to
fix the most obvious problem and either ignore the rest or ask people to
spend time reporting them separately. Therefore, I ask people to file
bugs separately in the first place in order to make our lives easier.

-- 
Colin Watson                                       [cjwatson at ubuntu.com]