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

Bug#706361: gti review



Hello,

Thank you very much for the detailed review of the package. I've taken some
time to address your feedback, as well as share some improvements upstream.

You can get the latest version of the package with
  dget -x http://mentors.debian.net/debian/pool/main/g/gti/gti_1.1.0-1.dsc
(note the version change)

Or you can browse the packaging work repository at
  https://gitorious.org/deb-packaging/gti-deb-packaging
or
  https://github.com/felixc/gti-deb-packaging

Some inline point-by-point responses to your comments:

> There is unnecessary lintian-override for "no-upstream-changelog".
> Please remove it.

Removed. I take it it is not customary to provide overrides for "pedantic"
warnings?

> By the way why you didn't forward man page upstream yet?

No reason, I was just focusing on getting the packaging work done. I have
now forwarded the man page upstream, where it has been accepted. The updated
package now reflects that.

> Package version looks a little bit unconventional. Although it's not a
> problem I would suggest to use just version "1.0.4-1" and backport
> patches from unreleased upstream commits if necessary.

Yes, I was very uncertain about the version number. I consulted with an
acquaintance who is a Debian Maintainer, and he suggested the format
incorporating the git commit hash and date. In any case, upstream has bumped
the version number, and I am now using the new version directly, so the
problem has gone away.

> ## debian/watch

Thanks to the new upstream version, debian/watch now works properly.

> debian/rules contains unnecessary comments (dm-make template
> remnants?) -- please consider removing 'em.

Removed!

> To save some space you might want to compress .deb file with xz
> instead of gz.

Changed! Thanks for this info, I hadn't seen any documentation on this
option, but it makes a lot of sense, so I'm glad to know about it now.

> License name "MIT" is incorrect even though upstream may refer to this
> license as such. "MIT" is considered ambiguous by the Free Software
> Foundation.

Yes, I saw that, which is why I included the full license text. I assumed
that would disambiguate it. Does the "ambiguous" designation mean that the
name "MIT" should simply not be used? I have changed it to "MIT Old Style".

> Priority "extra" is probably better changed to "optional"

Good point; I re-read the policy details on priorities and now agree with
the "optional" level.

> IMHO "Suggests: sl" is incorrect because it suggests unrelated package.
> I would replace it with "Enhances: git".

I again went back to re-read the documentation on this, and now can see
why Enhances is the better choice. Updated.

> Versioned build-depends on "dpkg-dev (>= 1.16.1~)" is unnecessary and
> may be completely dropped.

I suspected this might be the case, but wasn't sure. Removed.

> Specifically I think it is a good practice to maintain "Forwarded" and
> "Last-Update" headers.

All patches now have these.

> debhelper 9 in compat 9 mode automatically export build flags which
> makes unnecessary to include "/usr/share/dpkg/buildflags.mk" or set
> "DPKG_EXPORT_BUILDFLAGS" as we used to do with dh 8 to import
> hardening flags.

Oh, that's excellent. It certainly makes the Makefile much cleaner. I have
also forwarded the relevant Makefile changes upstream.

Again, thank you for your review. I appreciate all the information you've
shared, and I think the package is much improved after these changes.

All the best,

Felix

Attachment: signature.asc
Description: Digital signature


Reply to: