* Tue Dec 27 2005 Enrico Scholz <enrico.scholz@informatik.tu-chemnitz.de> - 1.10-0.7
- added initng initscripts
- do not make the 'fnord' user a member of the 'www' group because
  this group does not exist in a minimal installation
- added /srv/www/fnord directory
- apply -setgid patch


http://ensc.de/fedora/fnord.spec
http://ensc.de/fedora/fnord-1.10-0.7.src.rpm

Hi, Enrico. Your package looks very well. I just have a couple of questions
before I approve it;

1.) You provide initng startup script, but what about classical LSB init
scripts? Please refer to existing init scripts or the following document
in case you need some aid writing them: [1]

http://www.linux-foundation.org/spec//booksets/LSB-Core-generic/LSB-Core-generic.html#TOCSYSINIT

2.) Apart from init scripts -- is fnord able to run only in standalone mode
or also from inetd? In case it is able to run from inetd, please provide
a xinetd file for the server.

3.) Why do you link against dietlibc, not glibc? Does dietlibc make sense for
anything apart from embedded environments? As far as I know, glibc can not
be easily removed from Fedora, nor am I able to imagine why would someone do
that.

Some additional comments and complete review:

* rpmlint:
E: fnord non-standard-gid /srv/www/fnord fnord
E: fnord non-standard-dir-perm /srv/www/fnord 0750

The ownership and path are both fine.
But could you please pick some more standard path than /srv?
Would /var/www be a good choice?

E: fnord statically-linked-binary /usr/sbin/fnord-httpd
E: fnord statically-linked-binary /usr/sbin/fnord-idx
E: fnord statically-linked-binary /usr/sbin/fnord-cgi

I asked about linking against dietlibc in the comment above.

* The package is named according to guidelines
* Spec file name is fine
* Package meets the guidelines
* Package is licensed under GPL
* The license text is included in documentation
* To the extend I understand, the spec file is in American English
* The spec file is clear, legible and easily understandable
* The source matches upstream
     fnord-1.10.tar.bz2 = MD5(4c7d9f0e2b2f071d4687688f3018ba91)

What is the Source1: http://www.fefe.de/fnord/%name-%version.tar.bz2.sig good for?

* Tried compiling and running on i386 and x86_64 successfully
* Dependency list seems to be complete
* Package makes no use of locales
* Does not provide dynamically loaded libraries
* Not relocatable
* Package owns the directory it creates (though correctness of the path is
questionable, see comment at the top of this comment)
* Contains no duplicate entries in %files
* %files sections are fine and contain %defattr
* Contains proper %clean section
* Consistently uses macros
* Contains permissable content (code)
* No large quantities of documentation
* %doc files are not required for correct function
* No header files
* No static libraries
* No pkgconfig files
* No library files
* No devel subpackage
* No libtool archives
* No GUI
* No confilcts about files with anny other package
* %install begins with removal of PRM_BUILD_ROOT as it should
* All filenames are 7bit ASCII, so also valid UTF-8

#176582#c6 says that
> FESCo voted against allowing ipsvd to link statically against dietlibc in
> Fedora. I don't see how the outcome could be different for this package, but if
> you want it to also be voted upon, please say so.

Same for this package.

Packages depending on this were revoked, so I am refusing also this one.
If you feel that it was an incorrect decision, please reopen and let me know.

Also, if you would use the same reasoning as in #176582#c2, please do not
bother to reopen. Arguments like "are implemented correctly" are not only just
not good enough FESCo, but for everyone.