Intrepid fails LSB 3.2 tests on foomatic-filters

Due to the many regression bugs which the foomatic-filters package in Intrepid has (see changelog below) I propose to replace Intrepid's foomatic-filters by the final 4.0.0 release (which is currently in Jaunty).

An SRU request is uploaded to -proposed. Here is the debdiff

http://launchpadlibrarian.net/21399871/foomatic-filters_4.0.0%7Ebzr177-0ubuntu1_4.0.0-0ubuntu0.1.debdiff

Note that its is long, as due to the final release also the documentation is updated and the machine-generated files of the build system got regenerated. The following diff contains only the changes in active code (which actually modify the /usr/bin/foomatic-rip executable):

http://launchpadlibrarian.net/21399918/foomatic-4-intrepid-foomatic-4-final-changes-in-active-code.diff

The following debian/changelog contains all fixes on active code, the upstream ChangeLog I do not show here as it is much longer due to the documentation changes and the 4.0.0 final release.

foomatic-filters (4.0.0-0ubuntu0.1) intrepid-proposed; urgency=low

  * New upstream release (4.0.0 final release, BZR rev 195, all the changes
    against rev 177 as shipped by Intrepid are regressions against Hardy)
     - Let custom JCL options get correctly inserted.
     - Treat "None" as empty string in string and password options.
     - String and password option settings got inserted with the
       "Custom." prefix.
     - Fixed segfault which occured when using string options (LP: #318614).
     - Make foomatic-rip working correctly with PostScript of StarOffice 8.
     - The PostScript code of PostScript options did not get inserted when
       the spooler is CUPS and foomatic-rip had to convert incoming PDF to
       PostScript (LP: #299918).
     - Made merging of PJL options of the PPD and of the driver correctly
       working (LP: #318818).
     - Corrected the insertion of the "%%PageSetup" sections (LP: #318818).
     - Inserted PostScript option settings from pstops were not corrected
       with custom option settings done on the command line (numerical,
       string, password).
     - If there was only one PJL option in the PPD no PJL got added at all
       (LP: #303691).
     - Made foomatic-rip conforming with the LSB compliance tests (LSB bug
       #2423, LP: #318818).
     - Fixed infinite loop when page-specific option settings are used
       (LP: #318816).
     - Made foomatic-rip correctly working with custom JCL options.
     - Option setting insertions got corrupted when there is a composite
       option in the PPD file (Upstream bug #173, comment #23).

 -- Till Kamppeter <email address hidden> Mon, 19 Jan 2009 16:42:49 +0100

On Mon, Jan 19, 2009 at 04:12:02PM -0000, Till Kamppeter wrote:
> Due to the many regression bugs which the foomatic-filters package in
> Intrepid has (see changelog below) I propose to replace Intrepid's
> foomatic-filters by the final 4.0.0 release (which is currently in
> Jaunty).

This does not fit the SRU process as described at
<https://wiki.ubuntu.com/StableReleaseUpdates>. Micro-release exceptions
for SRUs must be approved by the TB, based on the presence of an upstream QA
process that provides a reasonable guarantee that the new release will be
regression-free relative to what we released with. Unless a micro-release
exception is first established, we should only be accepting targeted changes
that each qualify under the SRU policy.

Which of these changes do you believe meet the SRU process's requirement to
only fix high-impact bugs? I will say that I don't think LSB test failures
are themselves an SRU justification.

--
Steve Langasek Give me a lever long enough and a Free OS
Debian Developer to set it on, and I can move the world.
Ubuntu Developer http://www.debian.org/
<email address hidden> <email address hidden>

Steve Langasek [2009-01-20 6:50 -0000]:
> This does not fit the SRU process as described at
> <https://wiki.ubuntu.com/StableReleaseUpdates>. Micro-release exceptions
> for SRUs must be approved by the TB, based on the presence of an upstream QA
> process that provides a reasonable guarantee that the new release will be
> regression-free relative to what we released with.

Please note that this is not a general request to put in new
microreleases. The complete diff was provided, and we do not plan to
put in further upstream updates. This is just a bunch (some 10)
individual commits, and I don't particularly care whether they are
called '4.0 final' or 'merge fixes from upstream'.

> Which of these changes do you believe meet the SRU process's requirement to
> only fix high-impact bugs? I will say that I don't think LSB test failures
> are themselves an SRU justification.

That was my concern as well. The other fixes are regressions from
hardy, and thus acceptable in an SRU. How much of the diff is related
solely to LSB compatibility, and would disappear if that was taken
out?

The fixes which I did to make it LSB-compliant are also regressions against hardy, most prominently the security bug 318816. They also include that things like wrong treatment of supplying string or numerical option settings which are out of range and that PJL option settings do not get correctly inserted into the output data stream (which cuses also bug 303691). All problems described in http://bugs.linuxbase.org/show_bug.cgi?id=2423 are also failures of the basic functionality and regressions against Hardy (Hardy shipped foomatic-filters 3.0.x).

Accepted foomatic-filters into intrepid-proposed, please test and give feedback here. Please see https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you in advance!

There is an updated version which fixes a regression:

 foomatic-filters (4.0.0-0ubuntu0.2) intrepid-proposed; urgency=low
 .
   * debian/patches/foomatic-rip-segfault-on-jcl-merging.patch: Fixed
     segfault of the JCL merging process on long JCL command lists
     (LP: #321164). This was also reported on the Intrepid SRU
     4.0.0-0ubuntu0.1 in LP: #318614.

Can you please test this version instead/again, and report back here? Thank you!