reverse-build-depends matches by substrings

This bug was fixed in the package ubuntu-dev-tools - 0.45

---------------
ubuntu-dev-tools (0.45) intrepid; urgency=low

  [ Siegfried-Angel Gevatter Pujals ]
  * common.py:
     - Trying to read from a locked sqlite cookie database isn't a fatal
       error anymore.

  [ Adrien Cunin ]
  * update-maintainer:
     - check at the beginning of the script that the necessary files are
       readable/writable, and note which control files we are going to modify
     - at the end, only modify those control files, so that the script doesn't
       return 1 anymore when it was actually successful
  * pbuilder-dist:
     - Eliminated some warning with a better check for whether a given distro
       already has a pbuilder chroot in $BASE_DIR, when that distro is not
       known by the script
     - Added intrepid as a known distro
  * Return to previous versioning, without the ubuntu1 bit

  [ Jonathan Patrick Davies ]
  * buildd: Revert arch:status string format.

  [ Cesare Tirabassi ]
  * reverse-build-depends:
    - add -X option to grep-dctrl so that it only works with exact matches
      (LP: #272273).

 -- Adrien Cunin <email address hidden> Wed, 24 Sep 2008 16:01:09 +0200

I'm reverting this change, apparently dpkg-dctrl definition of full match is for the WHOLE field, so single package names will not be grepped. Perhaps one better solution could be to specify the pattern as a regex, for instance in between word boundaries (which means one would have to escape all possible regex special characters in the package name). I'm leaving this open anyway, perhaps reverse-build-depends should simply be removed since it doesn't do what it should do (and its anyway easier to just use dpkg-ctrl directly).

This perl script seems to be doing the job (output is print ala apt-cache).

(There's also http://people.ubuntu.com/~pitti/scripts/checkrdepends, but it needs a fast internet connection).

Cesare: Do you know the author / licensing / etc for the script you attached?

Its just something I wrote, I could add a license header if you think this could be useful.

The advantage is also that its pretty faster and works for multiple packages.

There is a better script on devscripts, called "build-rdeps", so we should switch to it. However, it needs some work so that it works in Ubuntu, which is: making it recognize restricted, universe and multiverse, and changing the default distribution to Ubuntu's development release (instead of "unstable").

Sieg's suggestions have been implemented by Ryan in lp:ubuntu-dev-tools (revno 367).

doc/reverse-build-depends.1 needs to be dropped and a manpage with the
text from ubuild-deps (but changing build-deps to ubuild-deps) added.

Further, why don't we keep the name reverse-build-deps so that people
who used it don't need to get used to a new (less intuitive) name?

All fixed in Bazaar (revno 368).

Great :).

(Now if only it used pod2man or something to generate the manpage from
the source instead of having that gibberish there... *g*).

This bug was fixed in the package ubuntu-dev-tools - 0.72

---------------
ubuntu-dev-tools (0.72) jaunty; urgency=low

  [ Jonathan Davies ]
  * README.updates: Added - lists steps to take when updating this package.
  * grab-merge: Added --help option and manpage (LP: #349109).

  [ Siegfried-Angel Gevatter Pujals ]
  * pbuilder-dist:
     - Add squeeze as a Debian distribution. Thanks to Marco Rodrigues.

  [ Nathan Handler ]
  * pull-debian-source:
    - Add support for etch/oldstable
    - Make script work for codenames (etch, lenny, squeeze, sid)

  [ Ryan Kavanagh ]
  * Ported devscripts' build-rdeps to Ubuntu and replaced
    reverse-build-depends. Updated it's manpage. (LP: #272273)

  [ Kees Cook ]
  * mk-sbuild-lv:
    - Fully handle missing build log directories (LP: #342154).
    - More generalized approach to Distro-specific logic (LP: #342158).

  [ Scott Kitterman ]
  * dgetlp:
    - Port to hashlib module instead of md5 (deprecated in Python 2.6)
  * Bump minimum python-all-dev version to 2.5

 -- Scott Kitterman <email address hidden> Wed, 15 Apr 2009 22:51:14 -0400