[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#456802: lintian: "strong implies weak" dependency check broken



Julien Cristau <jcristau@debian.org> writes:
> On Mon, Dec 17, 2007 at 17:55:02 -0800, Russ Allbery wrote:

>> It looks like this should only trigger when you're listing the exact
>> same substvar in both depends and recommends, which seems to be exactly
>> the problem that this check is warning about since that will generate
>> the same results in both Depends and Recommends.  Am I misunderstanding
>> something?

> The xserver-xorg stanza in debian/control looks like this:
> Package: xserver-xorg
> [...]
> Depends: xserver-xorg-core (>= 2:1.4-3), xserver-xorg-video-all | xserver-xorg-video-2, xserver-xorg-input-all | xserver-xorg-input-2, ${misc:Depends}, xkb-data | xkb-data-legacy, x11-xkb-utils
> Recommends: libgl1-mesa-dri, udev, discover1 | discover, ${F:XServer-Xorg-Detect-Depends}
> [...]
>
> I understand the intent of this check, but it doesn't seem to actually
> work here.

Oh, I see what the problem is.  This is the peril of reusing a generic
dependency parser that has its own ideas of what a valid package name is.
It's actually reducing both substitution variables to essentially the
empty string and then comparing them as equal because they don't look like
package names.  Bleh.

I'll figure out some way of fixing the false positive for the next
release, probably by transforming the substitution variables into
something that the dependency parser will swallow.

Sorry about that.

-- 
Russ Allbery (rra@debian.org)               <http://www.eyrie.org/~eagle/>



Reply to: