Bug#834768: RFS: codicefiscale/0.9-1
On 2016-08-18 at 22:27:42 +0000, Mattia Rizzolo wrote:
> FYI: no need to explicitly CC d-mentors@, RFSes are somehow sent there
> anyway.
ack
> This is DPMT, where the usage of git is mandated, so I expect the git
> repository to match the generated .dsc (hence I'm ignoring mentors here)
it does (hopefully) match, yes
> A few small things I'd like to see:
>
> * you email address in d/copyright
added in git
> * Files-Excluded in d/copyright doesn't list all the files that are
> removed (at least according to `git diff --stat
> upstream/0.9..upstream/0.9+ds0`)
> besides, wrapping that list might not be a bad idea
Uhm, I used uscan to remove the files, so nothing that wasn't listed was
removed.
Do you mean that I should explicitely list all of the content of the
directory ``codicefiscale.egg-info``, instead of just listing the
directory?
> * Also would be nice to see Build-Depends wrap-and-sort'ed
done in git
> * python3-codicefiscale uses ${python:Depends} instead of
> ${python3:Depends}
uooops, fixed in git
> * why do you disable the tests? (a comment on d/rules might not be a
> bad idea here either)
> + I see setup.py lists non-existant tests, if that's the issue maybe
> you can get that tests= arg removed (or the actual tests included)
> upstream?
That's exactly the issue, I've added a comment with a pointer to
https://github.com/ema/pycodicefiscale/issues/6
> * in d/watch, you dversionmangle '.ds0' away, but you're using '+ds0'
> actually, so it's not actually mangling anything
I hate single character typos, fixed in git (it appeared to work in
practice, but only because of versions ordering)
> * just quickly skimming over the README, I think it would make sense to
> include in the binaries, as it provides quick documentation (I think)
yes, it does, you're right (added in git)
--
Elena ``of Valhalla''
Reply to: