On Sat, 2003-02-15 at 06:42, Jay DeKing wrote:
> Not sure what you mean by "non-tree leaves," but glad to have some 
> confirmation of the rpmdrake hangs.
> 
> On another note, every package I update with urpmi recently has been reported 
> as having "bad signatures." None of these are Contribs - all standard Cooker.

As a confirmation I'm getting a number of them as well but not
everything.  (most have been reported through bugzilla.)
> 
> And, yesterday I found that urpmi was telling me that everything was already 
> install, though it obviously wasn't when I went to the mirror that I 
> reference. I had run urpmi.update already (and did it again after getting 
> that message), and the synthesis.hdlist.cz was the newest file in the mirror 
> (so that wasn't out-of-sync either).
> 
> I'll try again today when I get home from work.
> 
> Thanks
> Jay
> 
> On Friday 14 February 2003 02:31 pm, Guillaume Cottenceau honored me with this 
> communique:
> > Jay DeKing <[EMAIL PROTECTED]> writes:
> >
> > [...]
> >
> > > Yes, I've had this problem for at least two versions - so bad, in fact,
> > > that I've been doing all of my updates from the command line via urpmi. I
> > > open Konqueror on my favorite mirror and sort by date, rpm -q for the
> > > newest packages to see if I need to update them, and run urpmi for them
> > > if necessary.
> >
> > As for me, it seems that it happens when non tree leaves are
> > selected. Selecting another thing seems to workaround.
> >
> > > rpmdrake, on the other hand, hangs completely (not every time, but if not
> > > on the first package group, soon thereafter) if I use it to install
> > > updates. It appears to install the software, but does not return the "All
> > > packages were installed successfully" dialog.
> >
> > I think I had seen that at the end of normal rpmdrake too.
> >
> > > Running gps or ps -ef shows that grpmi (called by rpmdrake) has zombied.
> > > Worse, it won't respond to a hangup signal and can't be killed, except by
> > > sending rpmdrake a hangup signal.
> > >
> > > The package being installed when the hangup occurs *appears* to have gone
> > > in successfully, but without confirmation via completion of the rpmdrake
> > > process, how can I be certain? "rpm -qa <packagename>" says so, but I
> > > doubt that this checks every facet of the new software. What if something
> > > is missing from some critical packages?
> >
> > rpm -V?
> >
> > > I've checked all the dependencies of rpmdrake and grpmi and made sure
> > > that they were updated.
> >
> > Yes, the problem is still there. I've yet to investigate & try to
> > fix or work-around the problem.


Reply via email to