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

Re: RFR: blogofile -- Static website compiler and blog engine



Andreas Rütten <AndreasRuetten@gmx.de> writes:

> this is a RFR (Request For Review) for my blogofile package. 

Thank you for working on this package, I am looking forward to its
addition to Debian.

> - Since a few days the watch file included isn't working. Probably this
>   is because of a change of the GitHub frontend on October 12th (see
>   https://github.com/blog/958-the-code-tab). I informed Gunnar Wolf
>   about this because he is the maintainer of githubredir.d.n.

I don't understand the redirector. I also don't use Github.

Here is a watchfile that worked for me to get the latest tag as a
tarball:

    # watch control file for uscan
    version=3

    opts=filenamemangle=s/.+\/tarball\/(.+)$/blogofile-$1\.tar\.gz/ \
        https://github.com/EnigmaCurry/blogofile/tags/ /EnigmaCurry/blogofile/tarball/(.+)

Someone else here may suggest a better configuration. Or maybe the
redirector will be fixed soon.

> - There is no upstream changelog file. Lintian complains about it. 
>    - Is it right, that a missing upstream changelog file is not a
>      strict must have and it will not preventing the package to be
>      included into Debian?

Correct. You should explicitly acknowledge the lack of a changelog by
adding a Lintian override for that warning. Explain why in a comment to
the override.

> - Lintian complains about an extra-license-file. I included this file
>   because the upstream LICENSE.txt contains beside the actual Expat
>   (MIT) license some thoughts from the Author about his view on the
>   license an Free Software in general.
>    - Should it be included or removed from the package?

Good question. Sadly the author has decided to use that file for some
non-license information. You could make a request to the upstream
developers to put that non-license information in a separate file.

> - The package builds fine with debuild respectively dpkg-buildpacke but
>   not with pbuilder or cowbuilder. That's sound for me that there are
>   missing build dependencies but unfortunately I couldn't identify
>   which one until now.

The failures should tell you what the dependencies are. You might want
to discuss them on the Python forum for Debian.

> - My resulting package contains a blogofile.egg-info directory. I'm not
>   sure if it's needed for the runtime of blogofile or only during the
>   build of the package. I ask on #debian-mentors and got that there is
>   no clear answer to this. Sometimes a package need the .egg-info
>   directory at runtime and sometimes not.
>    - Do you know if it's needed or not?
>    - Can somebody explain my how to identify this point?

Ask that question on the Debian Python forum also. There is a standard
way to instruct the build system to install the package Debian-friendly.

> - Suggest or recommend a VCS, or explicitly git?
>   Blogofile itself will greatly work without any Version Control
>   System, but there are these two points:

I think you should not favour any particular VCS.

Since you say the package is useable without any VCS, there should be
none in the Depends field. The Recommends field should list them all as
alternatives (so it is satisfied by any one of them), and the
Suggests field should list all of them together.

> - What do you think about shipping the example site as .zip file in the
>   Debian package?

Perhaps unpack it, and install it as a separate binary package (maybe
‘blogofile-examples’).

> - Should they go into a separate -data package?

Not that name, that would imply that it's *necessary* data for the
package, which is not true.

> Thank you for keep reading up to here. Again any review, comment, help
> or hints are really welcome.

Thanks for your perseverence.

-- 
 \     “My house is on the median strip of a highway. You don't really |
  `\        notice, except I have to leave the driveway doing 60 MPH.” |
_o__)                                                   —Steven Wright |
Ben Finney

Attachment: pgpeALhSqZSLw.pgp
Description: PGP signature


Reply to: