Re: Deficiencies in Debian
On Mon, Sep 13, 1999 at 03:55:14PM +0200, Hartmut Koptein wrote:
> > > This is silly and is never going to happen. People keep uploading new
> > > packages with RCB's, and the RCB list keeps growing, regardless of how
> > > many RCB's are closed. That _will not_ change until potato freezes.
> > Then we stop uploading if we reach 50 RCB. Esay, ehh? :-)
> > Maybe if we had a better definition of which bugs are release critical
> > bugs and which are package critical bugs. (E.g., xemacs21, which has
> > never gotten out of the early release test phase, has several "release
> > critical" bugs.) I've asked before if someone (this would probably have
> > to be the release manager) would do a list of the bugs that are really
> > "release critical" but people seem more interested in bemoaning the size
> > of the RCB list than in whittling it down to reasonable proportions.
> Every package that doesn't compile is release critical. If it does compile
> but is isn't useful it is important (not installable, seg-fault, ...).
Release critical implies "we can't release until this bug is fixed."
This is apparantly the interpretation used by those who say we can't
freeze until the bug list is reduced. For some set of packages that
simply isn't true--we can remove the package if no one cares enough to
fix the bug. My point is that we need a list of the bugs that actually
hold up the release. I'm willing to spend time fixing bugs, but I've
found that I spend as much time wading through the "release critical"
bug list _looking_ for worthwhile bugs as I actually spend working on
them. It is not reasonable to ask people to fix bugs without giving some
direction as to where their effort is actually needed. The latest
debian-qa proposal may finally address this.