[Lsb-infrastructure] [Bug 1875] Build environment RPM DB contains errors

[bugzilla-daemon at linux-foundation.org]

http://bugs.linuxbase.org/show_bug.cgi?id=1875


Mats Wichmann <mats at freestandards.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|lsb-buildenv                |lsbpkgchk
            Product|Development Environment     |Application Tests
   Target Milestone|---                         |4.0




--- Comment #7 from Mats Wichmann <mats at freestandards.org>  2008-02-04 15:08:56 ---
Basically, a couple of well-known problems here.

SIGTAG_MD5 calculated value doesn't match expected value

lsbpkgchk problem.


Error: Didn't see expected dependency lsb=3.1

Developer error, didn't put the mandatory Requires: lsb  (someversion) in his
specfile.


Error: checkRpmIdx() unexpected Index tag=1140 type=4 offset=2114 count=46

yeah, lsbpkgchk only knows about the things it was coded to know about, and
those are driven from what the spec has innit - which has we know is an
extremely old version based on the fact that when rpm's format was being spec'd
one particular distro had not forklifted themselves off of rpmv3.


checkRpmArchiveFilename: file opt/SpliX not FHS compliant

lsbpkgchk can't guess the "registered LANANA name" (assuming you can get one,
Hi, Russ :-)  so you have to tell it what that is.  Once you tell it that (via
the "-L lananame" commandline option), it checks that pathnames conform to the
guideline /opt/lanananame, /etc/opt/lanananame, /var/opt/lanananame only as
being legal paths.  I would suspect that if there were package relocation in
play, lsbpkgchk would get it wrong.  


So nothing here is a a problem with the buildenv, it's basically lsbpkgchk
problems (which spec shortcomings leading to some of those).  Categorizing over
to lsbpkgchk.


-- 
Configure bugmail: http://bugs.linuxbase.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.