Bug 225906

I'd be happy to review this package. Look for a full review in a bit. 

OK - Package meets naming and packaging guidelines
OK - Spec file matches base package name.
OK - Spec has consistant macro usage.
OK - Meets Packaging Guidelines.
OK - License (GPL)
OK - License field in spec matches
OK - License file included in package
OK - Spec in American English
OK - Spec is legible.
OK - Sources match upstream md5sum:
dd965bdacbb86ce2a6498829fddda6b7  iptables-1.3.7.tar.bz2
dd965bdacbb86ce2a6498829fddda6b7  iptables-1.3.7.tar.bz2.1
See below - BuildRequires correct
See below - Package is relocatable and has a reason to be.
See below - Package has %defattr and permissions on files is good.
OK - Package has a correct %clean section.
See below - Package has correct buildroot
OK - Package is code or permissible content.
OK - Packages %doc files don't affect runtime.

OK - Headers/static libs in -devel subpackage.
See below - Spec has needed ldconfig in post and postun
See below - .so files in -devel subpackage.
See below - -devel package Requires: %{name} = %{version}-%{release}

OK - Package compiles and builds on at least one arch.
OK - Package has no duplicate files in %files.
OK - Package doesn't own any directories other packages own.
OK - Package owns all the directories it creates.
See below - No rpmlint output.
OK - final provides and requires are sane:

SHOULD Items:

OK - Should build in mock.
OK - Should build on all supported archs
OK - Should function as described.
See below - Should have subpackages require base package with fully versioned
depend.
See below - Should have dist tag
OK - Should package latest version
21 outstanding bugs - check for outstanding bugs on package.

Issues:

1. The Source URL is no longer correct.
Suggest: http://www.netfilter.org/projects/iptables/files/iptables-1.3.7.tar.bz2

2. Can the
Prefix: %{_prefix}
be removed? Is there any reason this package needs to be relocatable?

3. Minor: The BuildRequires: /usr/bin/perl
isn't required, as perl is part of the default build root.

4. Is the %defattr(-,root,root,0755) needed? or will %defattr(-,root,root,-) work?

5. Use the default correct buildroot:

      %{_tmppath}/%{name}-%{version}-%{release}-root-%(%{__id_u} -n)

6. Any reason the package makes a static lib instead of a shared lib?
Does anything use iptables-devel? Might be nice to remove the .a and
make a shared lib instead.

7. The devel and ipv6 subpackages have

Requires: %{name} = %{version}

Suggest: change that to:

Requires: %{name} = %{version}-%{release}

8. Our buddy rpmlint says:

a)
W: iptables summary-ended-with-dot Tools for managing Linux kernel packet
filtering capabilities.

Suggest: remove . at end of summary.

b)
E: iptables tag-not-utf8 %changelog

Looks like the checkins from
Trond Eivind Glomsr<F8>d <teg>
are not proper ut8f.

Suggest: Change to utf8.

c)
W: iptables strange-permission iptables.init 0755

Suggest: change the source file permissions to 644 and install it
as 755 only when installing?

d)
E: iptables non-utf8-spec-file iptables.spec

The encoding of the entire spec is not utf8.
Suggest: run iconv on it?

e)
E: iptables broken-syntax-in-scriptlet-requires Requires(post,postun): chkconfig

There was a rpm bug that made that syntax not work right.

Suggest: Change to:

Requires(post): /sbin/chkconfig
Requires(preun): /sbin/chkconfig

f)
W: iptables redundant-prefix-tag

Suggest: remove prefix.

g)
W: iptables rpm-buildroot-usage %prep rm -rf %{buildroot} 

Suggest: remove rm in prep stage. It's not needed here.

h) 
W: iptables macro-in-%changelog postun
W: iptables macro-in-%changelog preun

Suggest: Change any %macro names in changelog to be %%macro so rpm
doesn't try and expand them.

i)
E: iptables no-cleaning-of-buildroot %install

Suggest: add the rm -rf $RPM_BUILD_ROOT here.

j)
W: iptables conffile-without-noreplace-flag /etc/rc.d/init.d/iptables

Suggest: No need to mark the init script as a config file.

k) 
E: iptables non-readable /etc/sysconfig/iptables-config 0600

I guess this can be ignored. Not sure how much security it provides however.

l)
W: iptables service-default-enabled /etc/rc.d/init.d/iptables

Normally this is a no-no, but in this case I think we do want it
enabled by default. 

m)
W: iptables no-reload-entry /etc/rc.d/init.d/iptables

reload doesn't make sense here.

n)
W: iptables-devel spurious-executable-perm /usr/include/iptables.h
W: iptables-devel spurious-executable-perm /usr/include/ip6tables.h
W: iptables-devel spurious-executable-perm /usr/include/iptables_common.h

Suggest: Those headers should be mode 644. 

o)
E: iptables-ipv6 executable-marked-as-config-file /etc/rc.d/init.d/ip6tables

See item l above... should probibly not be a config file.

p)
W: iptables-ipv6 service-default-enabled /etc/rc.d/init.d/ip6tables
W: iptables-ipv6 no-reload-entry /etc/rc.d/init.d/ip6tables
W: iptables-ipv6 incoherent-init-script-name ip6tables

Ignore.  

9. Minor: Consider adding a dist tag?

10. 21 outstanding bugs. None appear to be directly packaging related.
However, there are some high priority ones and the secmark bug should
would be very nice to solve before f7.

Setting the review flag to ? here, since this is under review. 

(In reply to comment #2)
> 10. 21 outstanding bugs. None appear to be directly packaging related.

Bug 216733 is packaging related, if you count files in the tarball that should 
be installed by the rpm a packaging problem (I would).

*** Bug 238791 has been marked as a duplicate of this bug. ***

Hey Thomas, any chance we could move this forward? 
Or have you made changes already that you would like me to check here... 

I think I forgot to report back to this bugzilla.
Please have a look at the latest package in rawhide (1.4.0-X). I have made lots
of changes because of the review in version 1.3.8-2, already.

Sorry for the delay here. Looking much better now:


OK - Package meets naming and packaging guidelines
OK - Spec file matches base package name.
OK - Spec has consistant macro usage.
OK - Meets Packaging Guidelines.
See below - License
See below - License field in spec matches
See below - License file included in package
OK - Spec in American English
OK - Spec is legible.
OK - Sources match upstream md5sum:
90cfa8a554a29b0b859a625e701af2a7  iptables-1.4.0.tar.bz2
90cfa8a554a29b0b859a625e701af2a7  iptables-1.4.0.tar.bz2.orig
OK - BuildRequires correct
OK - Package has %defattr and permissions on files is good.
OK - Package has a correct %clean section.
OK - Package has correct buildroot
OK - Package is code or permissible content.
OK - Packages %doc files don't affect runtime.

OK - Headers/static libs in -devel subpackage.
See below - Spec has needed ldconfig in post and postun
See below - .so files in -devel subpackage.
OK - -devel package Requires: %{name} = %{version}-%{release}

OK - Package compiles and builds on at least one arch.
OK - Package has no duplicate files in %files.
OK - Package doesn't own any directories other packages own.
OK - Package owns all the directories it creates.
See below - No rpmlint output.
OK - final provides and requires are sane:

SHOULD Items:

OK - Should build in mock.
OK - Should build on all supported archs
OK - Should function as described.
OK - Should have subpackages require base package with fully versioned depend.
OK - Should have dist tag
OK - Should package latest version

Issues:

1. You have the License as GPLv2.
I see a mix in the source files of: GPL, GPLv2, GPLv2+, GPLv2 or GPLv3.
I think this results in: GPL+ for the license?

2. Any reason the package makes a static lib instead of a shared lib?
Does anything use iptables-devel? Might be nice to remove the .a and
make a shared lib instead.

3. rpmlint says:

iptables.src: W: strange-permission iptables.init 0755
iptables.x86_64: E: non-readable /etc/sysconfig/iptables-config 0600
iptables.x86_64: W: service-default-enabled /etc/rc.d/init.d/iptables
iptables.x86_64: W: service-default-enabled /etc/rc.d/init.d/iptables
iptables-ipv6.x86_64: E: non-readable /etc/sysconfig/ip6tables-config 0600
iptables-ipv6.x86_64: W: service-default-enabled /etc/rc.d/init.d/ip6tables
iptables-ipv6.x86_64: W: incoherent-init-script-name ip6tables

All look ignorable.

So, items 1 and 2 look to be the last issues to address... thoughts?

Any news here? 

Since it's been so long I would be happy to run through my checklist again against the current package... 

Or if you would care to look and address those last 2 items, I can look after that.

Please have a look at iptables-1.4.2-1 in rawhide.

1) License is GPL+, now
2) The old libraries are still static, the new ones are shared. Making the old ones static, could break build scripts and the old libraries should be gone in the near future.

I went ahead and ran through my checklist again: 

OK - Package meets naming and packaging guidelines
OK - Spec file matches base package name. 
See below - Spec has consistant macro usage. 
OK - Meets Packaging Guidelines. 
OK - License (GPL+)
OK - License field in spec matches
OK - License file included in package
OK - Spec in American English
OK - Spec is legible.
OK - Sources match upstream md5sum:
a138d1c2e74321e0e4e228a9fb301c9a  iptables-1.4.2.tar.bz2
a138d1c2e74321e0e4e228a9fb301c9a  iptables-1.4.2.tar.bz2.orig

OK - BuildRequires correct
OK - Package has %defattr and permissions on files is good. 
OK - Package has a correct %clean section. 
OK - Package has correct buildroot
OK - Package is code or permissible content. 
OK - Packages %doc files don't affect runtime. 
OK - Package has rm -rf RPM_BUILD_ROOT at top of %install

OK - Headers/static libs in -devel subpackage. 
OK - Spec has needed ldconfig in post and postun
OK - .pc files in -devel subpackage/requires pkgconfig
OK - .so files in -devel subpackage.
OK - -devel package Requires: %{name} = %{version}-%{release}
OK - .la files are removed. 

OK - Package compiles and builds on at least one arch. 
OK - Package has no duplicate files in %files. 
OK - Package doesn't own any directories other packages own. 
OK - Package owns all the directories it creates. 
OK - Package obey's FHS standard (except for 2 exceptions)
See below - No rpmlint output. 
OK - final provides and requires are sane.

SHOULD Items:

OK - Should build in mock. 
OK - Should build on all supported archs
OK - Should function as described. 
OK - Should have sane scriptlets. 
OK - Should have subpackages require base package with fully versioned depend. 
OK - Should have dist tag
OK - Should package latest version
OK - Should not use file requires outside of /etc, /bin, /sbin, /usr/bin, or /usr/sbin
See below - check for outstanding bugs on package (merge reviews/rename/re-reviews).  

Issues: 

1. MINOR: Can you stick with one of $RPM_BUILD_ROOT or %{buildroot} ?

2. rpmlint says: 

iptables.src: W: strange-permission iptables.init 0755
iptables.x86_64: E: non-readable /etc/sysconfig/iptables-config 0600
iptables.x86_64: W: service-default-enabled /etc/rc.d/init.d/iptables
iptables.x86_64: W: service-default-enabled /etc/rc.d/init.d/iptables
iptables-ipv6.x86_64: E: non-readable /etc/sysconfig/ip6tables-config 0600
iptables-ipv6.x86_64: W: service-default-enabled /etc/rc.d/init.d/ip6tables
iptables-ipv6.x86_64: W: service-default-enabled /etc/rc.d/init.d/ip6tables
iptables-ipv6.x86_64: W: incoherent-init-script-name ip6tables

All those can be ignored. 

3. You might consider using one of the standard ways to remove rpath:
http://fedoraproject.org/wiki/Packaging/Guidelines (ie, sed instead of perl)

4. There are a few packaging related bugs that would be nice to fix up: 
https://bugzilla.redhat.com/show_bug.cgi?id=462207
https://bugzilla.redhat.com/show_bug.cgi?id=432617

The consistent use of macros is a MUST, so if you can fix that up I can 
approve this review.

Oh. I also just noticed that this package doesn't seem to honor the standard rpm CFLAGS. Is there any reason for that?

Please have a look at iptables-1.4.2-3 in rawhide.

Solved Issues:
1) using %{buildroot} everywhere
3) using sed instead of perl
4) #462207 has already been addressed in iptables-1.4.2-1
   #432617 will require basic firewall changes and is therefore not possible now
5) using RPM_OPT_FLAGS, but with -fno-strict-aliasing (needed for libiptc)

Excellent. I don't see any further blockers, so this package is APPROVED. 

Thanks for fixing things up.