Re: Need "Deferred" status for bug system
On 8 Feb 1999, Adam Di Carlo wrote:
> On Sun, 7 Feb 1999 13:05:47 -0800 (PST), brian holgate <brian@ashlu.bc.ca> said:
> > The current practice of downgrading bugs in order to release leads
> > to large numbers of old bugs. I never file bug reports because there
> > are almost always at least 3 similar bug reports on file, all too
> > often classified as Normal regardless of how disruptive they are.
>
> I don't think anyone's ever downgraded bugs to get a release out. We
> downgrade bug severity when we feel they don't deserve the level of
> importance they have been assigned (hello, Santiago!). If you have
> counter examples, please cite bug#.
Hello Adam ;-)
Actually, I think we would need a new bug severity named "ugly", between
normal and important.
An ugly bug would be a bug that causes major annoyance, make a lot of
people to lose a lot of time, make people to ask questions over and over
again in debian-user, etc. but however, are very easy to fix.
Ugly bugs would not delay the release on an individual basis, but only
globally, i.e. we could decide that frozen will not become stable
until there are less than, say, 10 ugly bugs.
Examples of ugly bugs:
* The existence of a "cdrom" install method (does not make the system to
break, but may cause a lot of people to lose a lot of time).
* Packages which do not upgrade automatically due to name changes (causes
major annoyance).
* Packages which should replace some others but they fail to do so (dpkg
complains loudly about this even when using --force-overwrite). This
causes also major annoyance.
* standard or above packages still linked against ncurses3.4, which forces
a new install to have an obsolete library installed. This is obviosuly
ugly and trivial to fix.
The border between an ugly bug and a bug which makes the package to be
unsuitable for release is not always very clear, I admit to have reported
some bugs as of severity: important, that were later downgraded to normal
severity. The ugly severity would have been probably more appropriate
in this case.
The problem with not having such a severity level is that we probably
release slink as soon as there are not important or above bugs, but it
will still contain a lot of ugly bugs, which would have been trivial to
fix.
--
"da3fb72838db005ca6856ef2be27bc6a" (a truly random sig)
--
To UNSUBSCRIBE, email to debian-testing-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: