[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [The good, the bad and] The Ugly -- the cosmetic rant



Greetings.

It seems to me some people entirely miss the point. The issues I'm
talking about are not only those that I mentioned, maybe I'm not even
right in some point (though I doubt that - analogous to the
binary-indep comments - anyone would comment `main' like `This is the
entry point of the program'). Nevertheless, below is another few
examples, from two unnamed packages:

,----[ Diffstat ]
|  debian/changelog  |   45 ++++++++++++++++++++++++++++++++++++
|  debian/control    |   20 ++++++++++++++++
|  debian/copyright  |   19 +++++++++++++++
|  debian/dirs       |    1 
|  debian/docs       |    2 +
|  debian/pam        |    4 +++
|  debian/patch.diff |   14 +++++++++++
|  debian/postinst   |   24 +++++++++++++++++++
|  debian/postrm     |   23 ++++++++++++++++++
|  debian/prerm      |    8 ++++++
|  debian/rules      |   67 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
|  debian/test.sh    |   11 ++++++++
|  params.h          |    8 +++---
|  13 files changed, 242 insertions(+), 4 deletions(-)
`----

Notice the patch.diff and test.sh files... With a little attention,
those wouldn't have made it into the diff.gz.

,----[ debian/rules ]
| configure: configure-stamp
| configure-stamp:
| 	dh_testdir
| 	# Add here commands to configure the package.
| 	touch configure-stamp
`----

Well, configure is not a required target, it does nothing, and the
package in question is unlikely to get a configure script ever. Still,
the target is there. I've seen this in quite a lot of packages.

On the other hand, the same package contains this:

,----[ debian/rules ]
| # Build architecture-dependent files here.
| binary-arch: build install
| 	# Nothing to do.
`----

Now, that's something I consider a good comment. No `by default',
which suggests that there _is_ a way to make the inspected
debian/rules build binary-arch packages, it clearly says it doesn't do
anything.

To mention a bit more, there are README.Debian files, which aren't
properly filled out, or contain information that is already available
in the upstream docs, and is not Debian related, etc, etc, etc..


As to why do I post here instead of filing bugs with a patch? I
thought it might be useful to drop a note, so future packages will be
better.

(Oh, and the bug reports are progressing, I just need a little time to
verify that my scripts didn't report false positives)

Attachment: pgp39CamPjMJg.pgp
Description: PGP signature


Reply to: