On Mon, Oct 22, 2012 at 02:00:29PM +0200, Pavel Březina wrote:
On 10/19/2012 12:43 PM, Jakub Hrozek wrote:
>On Fri, Oct 19, 2012 at 10:19:25AM +0200, Pavel Březina wrote:
>>On 10/18/2012 03:24 PM, Simo Sorce wrote:
>>>On Thu, 2012-10-18 at 09:46 +0200, Pavel Březina wrote:
>>>>https://fedorahosted.org/sssd/ticket/1437
>>>
>>>Fedora F15 has been EOLed and will not get updates, why bother with
>>>F15 ?
>>>Also F15 systemd was quite immature, and at the time guidelines
>>>mentioned you shouldn't change from sysv to systemd after GA.
>>>Did you mean to put F16 there ?
>>>
>>>Simo.
>>>
>>
>>OK. New patch is attached.
>
Hi,
>> %if (0%{?enable_experimental} == 1)
>>@@ -57,6 +61,12 @@ Requires(post): initscripts chkconfig
>> Requires(preun): initscripts chkconfig
>> Requires(postun): initscripts chkconfig
>>
>>+%if (0%{?use_systemd} == 1)
>>+Requires(post): systemd-units
>>+Requires(preun): systemd-units
>>+Requires(postun): systemd-units
>>+%endif
>
>I think here ^^ should be an %else that would otherwise Require
>initscripts. I don't think we need them in a pure systemd-enabled
>package.
I'm no spec file expert, but everything is basically copy and paste
from our Fedora spec file so I assume it should be correct.
The Fedora spec file says:
Requires(post): systemd-units initscripts chkconfig
Right, I think the initscripts Requires is spurious, but we keep
chkconfig to migrate from sysv scripts (which Pavel just showed me is
forbidden by the Packaging Guidelines).