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

Re: pysolfc-cardsets review



Not sure why you took the thread off-list, so I bounced your mail to the
list and am replying there too.

On Mon, 2012-02-20 at 23:27 +0100, Bernhard Reiter wrote:

> pysol-cardsets now also has those, plus
>     role::dummy

No point in having anything other than role::dummy on pysol-cardsets,
please remove the other ones. Same applies for pysol/pysolfc.

> I sat over this one with the original uploader (Miriam) for, I think, a
> couple of hours, searching for cardset-mini's origin on the web -- to no
> avail. Upstream has COPYRIGHT files in all cardset-* directories stating
> copyright and license information, but for this one, there's only
> license information, no copyright.
> 
> We found some of the card images on other websites and were able to
> narrow down about when the cardset was included in pysolfc-cardsets, but
> didn't find any copyright information. I also posted a request in the
> pysolfc forum
> https://sourceforge.net/projects/pysolfc/forums/forum/503708/topic/4880735 and sent an email to the maintainer, but didn't receive any reply.
> 
> I offered to remove the cardset, but Miriam said she'd rather have it in
> the package 'cause she liked it and probably felt that this wasn't too
> much of a problem threshold-of-originality-wise.
> Adding comment to paragraph 55 and a lintian override okay?

Add a comment, but not a lintian override. I wonder how the license is
known but not the copyright holder.

> There isn't really much information on the cardset bundle, nor are there
> any changelogs for the individual cardsets. There are some sparse
> changelog like bits on the pysolfc website
> http://pysolfc.sourceforge.net/ but those are more for pysolfc (the
> game, not the cardsets). As for pysolfc-cardsets, one can only see the
> number of cardsets included in different versions. Adding a lintian
> override okay?

No, adding an override is not appropriate when lintian is correct, nor
for info/pedantic/experimental tags. Just ignore the warning and ask
upstream to create a changelog.

> Fixed by using xz compression only for pysolfc-cardsets, not
> pysol-cardsets (which is a dummy package for transition).

Good.

> Okay to leave? Can't think of much more information to add...
> Or should I add an override?

Overrides are not appropriate to add here. Take a look at
warzone2100-data for an example of how to do this. You could describe
all the individual cardsets or if there are a lot of them, just the
major theme categories they fall into.

> I'm currently removing it in the override_dh_auto_install section of 
> debian/rules. I can forward this upstream (not really affiliated with
> them, though).

Please instead fix the script that generates the orig.tar.gz (and make a
new tarball), no point having these files there either. Please do ask
upstream about removing it.

I'm aware that devref/policy say that tarballs should be as unchanged as
possible from upstream, but these files and the dupes make the tarball a
fair bit bigger than it could be.

In addition to the .xvpics directory I am also seeing some .thumbnails
directories that are installed in the package and that lintian missed.
Please investigate if these are needed by pysolfc. If not, find what
program produced them and then file a lintian bug asking them to detect
them like they do .xvpics. If they are unused, please remove them from
the orig.tar.gz too and talk to upstream about removing them.

> Added to the override_dh_auto_install section in debian/rules. I'll
> forward a request upstream.

Please instead fix the script that generates the orig.tar.gz (and make a
new tarball), you could potentially make it much smaller.

> I'll look into that, but that doesn't really affect pysolfc-cardsets,
> just pysolfc, right?

Right.

> Anyway, my pysolfc-cardsets fixes are in git, and a fixed source package
> is on mentors.debian.net now.

Thanks for your work and please make the changes I suggested above.

-- 
bye,
pabs

http://wiki.debian.org/PaulWise

Attachment: signature.asc
Description: This is a digitally signed message part


Reply to: