Re: Bug system severities
On Sep 06, joost witteveen wrote
>
> Adrian Bridgett wrote:
> > IMHO I don't like the way that Ian just added the "Severity" field - I
> > think that it should have been discussed on the list beforehand.
>
> Oh, come on! The only discussion we haven't had on debian-* yet is
> whether we should first have a formal debate that later outlines
> how we are going to discuss changes to be proposed by somebody.
I think the bug system is am important aspect of Debian and that major
changes to it need thinking though. One person's idea of what's best
often disagrees with other peoples. I think it reflects well on Debian
that people care about it rather than just saying "fine, do what you want
then".
I agree that discussing things amongst 300 people takes a long time, but
discussion does make people think about what they are doing and their
justification. Maybe there should be a small development group
that could just "run over" decisions that affect Debian significantly -
an email to five or ten of the "top" developers may result in some good
ideas that hadn't previously been thought of.
> _EVERY_ other discussion has been here, it's just that there's always
> somebody who "misses" it.
>
> As others have pointed out: there has been discussion about the severity
> of the bugs. And, as with all other discussions, it first attracted
> many people, expressing there views, and then the discussion died down,
> and nothing happened. The people that joined the discussion expressed
> many different views, but I think they all agreed on one point: something
> should happen. And I'm also rather sure everybody would agree that
> what Ian did was an improvement.
I completely agree that the "severity" field is needed. It's just that I
think that a _quick_ discussion of what values it could take would have
been nice. I think Ian has made a good choice with his first four
values - critical/grave/normal/wishlist.
However he then added "difficult" which as someone else has pointed out
is a severity of bug fixability rather than of buginess. This is the sort
of thing which IMHO needs discussing amongst a wider audience before being
added to the system.
Adrian
--
.signature in post
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-devel-request@lists.debian.org .
Trouble? e-mail to templin@bucknell.edu .
Reply to: