On Tue, 2003-02-25 at 00:59, Tom Diehl wrote:
> On 25 Feb 2003, seth vidal wrote:
>
> > On Tue, 2003-02-25 at 00:10, Tom Diehl wrote:
> > >
> > > I see "Installing: kernel, athlon" in the new file at line 5163. Is that
> > > what you were looking for??
> > >
> > Right but I don't see [Install: kernel]
>
> Neither do I. For that matter I do not see anything in square brackets.
>
in a -d10 all things print oddly, ignore them :)
> > also - could you run rpm -q kernel
> > and send me the output b/c I'm not seeing that release tag anywhere
> > other than the last line, and well, that's odd. it should be in there
> > _SOMEWHERE_ other than there.
>
> The odd release tag is from a custom built kernel we installed.
right - but its obviously in your rpmdb - so it should be showing up
when yum does getupdatedhdrlist
so its comparing stuff installed vs stuff available.
In your case it should look at the installed kernel and the installed
kernel should report as newer.
But, it outputs those comparisons in the compareEVR.
So I'm looking through them and I _should_ be seeing a comparison b/t
0, 2.4.18, 24.7.x.1.nano.tjc vs 0, 2.4.18, 24.7.x
but I never see that comparison - which makes me think something is
broken.
so my first thought is, I'm parsing the release somehow, but I'm not
parsing the release, I'm asking for it from the rpmdb and using it
verbatim. So it's not a bug in parsing somewhere.
the closest thing I do see is this:
0, 2.4.18, 24.7.x vs 0, 2.4.18, 24.7.x = 0
The one thing I could think of is that somehow rpm is returning newer in
BOTH cases.
ok - one more fun thing
rpm -q --qf "%{name} %{version} %{release} %{arch}\n" kernel
send me the output of that - b/c _somewhere_ something in this list of
kernels is returning the WRONG thing, and afaict it has to be coming
from the rpmdb.
-sv
