[opensuse-packaging] shlib-fixed-dependency check
Hi, I see in libv4l: libv4lconvert0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = 0.6.4 libv4l1-0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = 0.6.4 libv4l2-0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = 0.6.4 Your shared library package requires a fixed version of another package. The intention of the Shared Library Policy is to allow parallel installation of multiple versions of the same shared library, hard dependencies likely make that impossible. Please remove this dependency and instead move it to the runtime uses of your library. Is that check really correct? I propose to disable it for 11.3 and enable it again afterwards. What do you think? Andreas -- Andreas Jaeger, Program Manager openSUSE, aj@{novell.com,opensuse.org} Twitter: jaegerandi | Identica: jaegerandi SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) Maxfeldstr. 5, 90409 Nürnberg, Germany GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126 -- To unsubscribe, e-mail: opensuse-packaging+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-packaging+help@opensuse.org
On Sunday 02 May 2010, Andreas Jaeger wrote:
libv4lconvert0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = 0.6.4 libv4l1-0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = 0.6.4 libv4l2-0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = Is that check really correct?
Looks genuine to me. Why do you consider it a false positive?
I propose to disable it for 11.3 and enable it again afterwards. What do you think?
I'd rather want to see a full factory rebuild against the latest rpmlint version to see how many/which packages are affected and are really failing the build because of that. 440 is still way below 1000 (the threshold for a failed build) Greetings, Dirk -- To unsubscribe, e-mail: opensuse-packaging+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-packaging+help@opensuse.org
On Sunday 02 May 2010 20:42:35 Dirk Müller wrote:
On Sunday 02 May 2010, Andreas Jaeger wrote:
libv4lconvert0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = 0.6.4 libv4l1-0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = 0.6.4 libv4l2-0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = Is that check really correct?
440 is still way below 1000 (the threshold for a failed build)
Greetings, Dirk Dirk,
440 is indeed way below 1000, unfortunately this seems to be 3 packages for which it is reported. The total badness comes with that to a total of 1320 which is over the threshold, so v4l would result in a failed build. Greetings, Raymond -- To unsubscribe, e-mail: opensuse-packaging+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-packaging+help@opensuse.org
On Sun, May 02, 2010 at 08:42:35PM +0200, Dirk Müller wrote:
On Sunday 02 May 2010, Andreas Jaeger wrote:
libv4lconvert0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = 0.6.4 libv4l1-0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = 0.6.4 libv4l2-0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = Is that check really correct?
Looks genuine to me. Why do you consider it a false positive?
Its not, the main "libv4l" just contains plugins.
I propose to disable it for 11.3 and enable it again afterwards. What do you think?
I'd rather want to see a full factory rebuild against the latest rpmlint version to see how many/which packages are affected and are really failing the build because of that.
440 is still way below 1000 (the threshold for a failed build)
3*440. Ciao, Marcus -- To unsubscribe, e-mail: opensuse-packaging+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-packaging+help@opensuse.org
On Sunday 02 May 2010, Marcus Meissner wrote:
libv4lconvert0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = 0.6.4 libv4l1-0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = 0.6.4 libv4l2-0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = Is that check really correct? Looks genuine to me. Why do you consider it a false positive? Its not, the main "libv4l" just contains plugins.
does "it's not" mean that it is not a false positive? I agree with that :-) This is the classic example of a correctly identified dependency issue. the plugins are not packaged in a library-specific subdir, and therefore would conflict when trying to install two versions of the library in parallel. Thats exactly what the check is trying to achieve.
440 is still way below 1000 (the threshold for a failed build) 3*440.
Yep, and I hope the package maintainer doesn't consider building from seperate spec files a valid workaround ;-) Greetings, Dirk -- To unsubscribe, e-mail: opensuse-packaging+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-packaging+help@opensuse.org
Hi, On Mon, 3 May 2010, Dirk Müller wrote:
libv4lconvert0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = 0.6.4 libv4l1-0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = 0.6.4 libv4l2-0.x86_64: E: shlib-fixed-dependency (Badness: 440) libv4l = Is that check really correct? Looks genuine to me. Why do you consider it a false positive? Its not, the main "libv4l" just contains plugins.
does "it's not" mean that it is not a false positive? I agree with that :-)
This is the classic example of a correctly identified dependency issue. the plugins are not packaged in a library-specific subdir, and therefore would conflict when trying to install two versions of the library in parallel. Thats exactly what the check is trying to achieve.
FWIW I completely agree with Dirk. The only improvement that could be done is to not report the same wrong dependency three times, but instead count it only once (the error message should ideally still mention all three subpackages, though). Ciao, Michael.
participants (5)
-
Andreas Jaeger -
Dirk Müller -
Marcus Meissner -
Michael Matz -
Raymond Wooninck