Bug#279754: lintian: missing-depends-line test appears too general
Hi Frank,
sorry for the slow reply, this one slipped through the cracks
somehow.
On Wed, Nov 17, 2004 at 12:50:29AM +0100, Frank Lichtenheld wrote:
> On Fri, Nov 05, 2004 at 04:00:10PM +1030, Ron wrote:
> > Hi, I triggered this warning building a small package that consists of
> > a couple of binaries and a shared lib that nothing other than these
> > binaries has any reason to use at present. I know it is bad form to
> > jam all this in a single package as a general rule etc. but the test
> > still appears to be failing for reasons unrelated to any opinion on that.
>
> Can you make this package available somewhere so we could take a look at
> it (or at least tell us its name ;)? Does it trigger any other errors and
> or warnings?
It was a preliminary package for the wacom support I'm working on
and it 'outgrew' this problem pretty quickly. IIRC it was lintian
clean for everything else by the time I reported this though.
> But yeah, I can confirm from looking at the code that if the binaries
> really only need libs from within the package there could be spurios
> warnings. (As almost all binaries are linked against libc anyway, they
> should be nearly non-existant, though)
I don't expect this to be a common misfire, but its behaviour in
this edge case seemed provably wrong and so worth reporting.
If it can be fixed to the benefit of the rest of the code, that
is good, but I wouldn't do anything ugly just to cover it up.
A local lintian-override is probably as good a hack as any if we
just want to mask it for the benefit of a couple of pathological
packages at this stage.
cheers,
Ron
Reply to: