On Wed, 02 Jul 2008 02:11:14 +0300 Eddy Petrișor wrote:

> Hello,
> 
> I have seen this report and I think I know what was the purpose of the 
> behaviour seen here:

I don't think that what I detected was done on purpose: just take a
look at the code and at my patch.  Without my patch, less_found_at_all
is set to false and then conditionally set to false inside the
iteration loop.  I mean: what is the use of a boolean variable which is
always false?

> 
> Francesco Poli:
> > Actually, whenever the given version is between two versions
> > the bugreport is marked as found in, the bugreport is wrongly
> > ignored.
> 
> I was about to report that this exact behaviour does *not* happen (and I am 
> still not sure how you 
> got it).

Did you try to issue the commands I provided as examples, getting
different results?  I still get substantially the same results (bug
#468926 is ignored in the first apt-listbugs invocation, but not in the second 
one: it should not be ignored in either one).

> 
> Rationale: when apt-listbugs is called from apt, a bug which already affects 
> my version shouldn't 
> stop me from upgrading since that would always keep me in place

This is true, but unrelated.

When I use the apt-listbugs list command, I want to know which bugs
affect a package or a given version of a package, regardless of
which version of that package I have installed on my system (if at
all!).
Hence, bug #468926, which is marked as found in versions
audacious/1.4.6-2, and audacious/1.5.0-2 (and now is marked as fixed in
version audacious-plugins/1.5.1-1, without any bug reassignment, but
that's another story...), *does* affect both audacious/1.5.0-1 and
audacious/1.5.0-2: it should be listed in both my apt-listbugs list
invocations.

> 
> 
> 
> I think the fix should be the addition of an option "--newsince $PACK/$VER" 
> which should instruct 
> apt-listbugs to show *only* the bugs which are present in the given version, 
> but were not present in 
> $VER.

Some way to make apt-listbugs ignore bugs that affect package/version,
but are already present in package/installedversion, can really be
useful, but should be filed as a separate wishlist bug, IMHO. 


-- 
 http://frx.netsons.org/doc/index.html#nanodocs
 The nano-document series is here!
..................................................... Francesco Poli .
 GnuPG key fpr == C979 F34B 27CE 5CD8 DC12  31B5 78F4 279B DD6D FCF4

Attachment: pgp2p7Jtj3HJZ.pgp
Description: PGP signature

Reply via email to