--- Comment #11 from Brett Lentz <> ---
(In reply to Zbigniew Jędrzejewski-Szmek from comment #10)
> I'll take the review.

Great! Thanks!

> > %global sum Python library to load and dump configuration files in various 
> > formats
> Using normal Summary: and then %summary subsequently saves one line ;)

Not true. The summary is used in 3 places because of the python2 & python3
sub-packaged. The macro saves copy/pasting the same text in 3 places.  :)

> > %global debug_package %{nil}
> That looks suspicious. Why do you need this?
> Package builds fine without it.


> > %{__rm}
> Eh, using a macro here is entirely pointless. It just makes the commands
> harder to read (and longer). The guidelines say that macros should be used
> for some *directories*
> [], but even that
> makes little sense nowadays.


> >{name}
> I know people love macros, but this makes it impossible to just click on
> this and open it in a browser… It's a matter of preference, but I don't see
> the advantage of using a macro here.


> > Source0:        %{url}/archive/RELEASE-%{version}.tar.gz
> This should be ...RELEASE_{%version}...


> > %defattr(-,root,root,-)
> Not necessary in Fedora and somewhat recent RHEL.


> - Large documentation must go in a -doc subpackage. Large could be size
>   (~1MB) or number of files.
>   Note: Documentation size is 2662400 bytes in 126 files.
>   See:
> It'd be nice to split out a python-anyconfig-doc subpackage with the docs.


> During build I see the following error:
> >     import cbor
> > ImportError: No module named cbor
> Is some dependency missing?

There is support for a backend (cbor) that does not currently have a package in
Fedora. If it's okay with you, I'd prefer to not block this package on the one
missing backend.

This anyconfig package is a dependency of the package I'm ultimately looking to
get into Fedora (molecule). However, anyconfig's support for cbor is not a part
of my critical path.

I am willing to work on a python-cbor package after anyconfig is in Fedora, if
that works for you.

> and later:
> > of anyconfig.backend.toml.Parser._load_from_stream_fn:6: 
> > WARNING: Definition list ends without a blank line; unexpected unindent.
> > /builddir/build/BUILD/python-anyconfig-RELEASE_0.9.3/anyconfig/backend/
> >  of anyconfig.backend.xml._tweak_ns:4: WARNING: Field list ends without a 
> > blank line; unexpected unindent.
> > /builddir/build/BUILD/python-anyconfig-RELEASE_0.9.3/docs/api/anyconfig.cli.rst:4:
> >  WARNING: autodoc: failed to import module u'anyconfig.cli'; the following 
> > exception was raised:
> > Traceback (most recent call last):
> >   File "/usr/lib/python2.7/site-packages/sphinx/ext/", line 658, 
> > in import_object
> >     __import__(self.modname)
> >   File 
> > "/builddir/build/BUILD/python-anyconfig-RELEASE_0.9.3/anyconfig/", 
> > line 42, in <module>
> >     sys.stdout = codecs.getwriter(_ENCODING)(sys.stdout)
> >   File "/usr/lib64/python2.7/", line 1009, in getwriter
> >     return lookup(encoding).streamwriter
> > TypeError: lookup() argument 1 must be string, not None
> > done

This looks like a bug in the docs. I'll point it out to upstream.

> And now the hard part: what is the difference in behaviour or output between
> anyconfig-2 and anyconfig-3?

There is no difference, AFAICS.

> And also (a question for upstream): why is the executable called
> "anyconfig_cli" and not just "anyconfig"?

Fixed in the spec until it's fixed upstream.

I've updated the spec and srpm. Same URLs as in comment #7.

You are receiving this mail because:
You are on the CC list for the bug.
You are always notified about changes to this product and component
package-review mailing list --
To unsubscribe send an email to

Reply via email to