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

Bug#702588: RFS: rrep/1.3.5-1



On 03/18/2013 03:33 AM, Paul Wise wrote:
> On Sun, 2013-03-17 at 17:44 +0100, Arno Onken wrote:
> 
>> I improved the watch file so that more extensions are found:
>> http://sf.net/rrep/rrep-(.+)\.(?:zip|tgz|tbz|txz|(?:tar\.(?:gz|bz2|xz)))
>> Were you referring to that?
> 
> I was also referring to the version regex, see the fourth item:
> 
> http://wiki.debian.org/debian/watch#Common_mistakes

Done. Now using
http://sf.net/rrep/rrep-(\d\.\d\.\d)\.(?:zip|tgz|tbz|txz|(?:tar\.(?:gz|bz2|xz)))

>> Why would you suggest that? The dh_installinfo way seems to be the
>> standard way that is described in the Maintainers Guide. I would prefer
>> to stick to that.
> 
> The info page should also be installed on distributions not using
> debhelper to build packages (like Fedora or Gentoo). The best way to
> achieve that is to perform all non-distro-specific build and
> installation steps in the upstream build system and anything specific to
> Debian or our derivatives using debhelper commands. 

Done. The info page is now installed by the upstream build system
instead of debhelper.

>> copy-file.c is part of gnulib. I tried the Debian version of gnulib and
>> I tried the latest upstream gnulib. I get the warning in all cases. I am
>> reluctant to change any gnulib files.
> 
> Please report this bug to the upstream of gnulib.

Done. The warning was silenced upstream and the latest release of gnulib
is included in rrep.

>> I removed the redundant check in rrep.c. rpmatch.c and regcomp.c are
>> part of gnulib. See response to copy-file.c.
> 
> Same response.

All cppcheck errors but one were fixed upstream and fixes are included
in the latest rrep release. The one remaining error is in lib/opendir.c:

[lib/opendir.c:140]: (error) Mismatching allocation and deallocation: dirp

It is a false positive of ccpcheck.

>> This is the standard texinfo file. I am reluctant to change anything here.
> 
> Makes sense, maybe you could report that upstream?

I took a closer look at the warnings of lacheck and they seem to be way
too picky. I didn't report them upstream.

>> Sorry, I have no clue about the linker error of the clang compiler. This
>> should work with any sane compiler.
> 
> Me either, maybe ask on the clang mailing lists what might cause that.

Done. The error is fixed in the latest release of rrep: inline functions
are changed to static inline functions.

I uploaded version 1.3.6-1. to mentors.debian.net and will file a new
RFS bug report.

https://mentors.debian.net/package/rrep

Please note that contrary to the report on mentors.debian.net, the
latest watch file actually does work. http://qa.debian.org/ was down
during upload.


Reply to: