On Wed, Jun 30, 2010 at 5:32 PM, Thomas Bächler <thomas@xxxxxxxxxxxxx> wrote:
> Am 30.06.2010 23:58, schrieb Daenyth Blank:
>> On Wed, Jun 30, 2010 at 17:47, Victor Lowther <victor.lowther@xxxxxxxxx> wrote:
>>> ---
>>> rc.sysinit | 2 +-
>>> 1 files changed, 1 insertions(+), 1 deletions(-)
>>>
>>> diff --git a/rc.sysinit b/rc.sysinit
>>> index 319ea60..037a9f2 100755
>>> --- a/rc.sysinit
>>> +++ b/rc.sysinit
>>> @@ -277,7 +277,7 @@ stat_busy "Removing Leftover Files"
>>> /bin/rm -f /var/lock/* &>/dev/null
>>> /bin/rm -rf /tmp/* /tmp/.* &>/dev/null
>>> /bin/rm -f /forcefsck &>/dev/null
>>> -(cd /var/run && /usr/bin/find . ! -type d -exec /bin/rm -f -- {} \; )
>>> +(cd /var/run && /usr/bin/find . \! -type d -delete )
>>> : >| /var/run/utmp
>>> /bin/chmod 0664 /var/run/utmp
>>> # Keep {x,k,g}dm happy with xorg
>>> --
>>> 1.7.1
>>>
>>>
>>
>> I don't think this is portable though.
>>
>
> Portability is not an issue. The beauty about this is that we can simply
> assume we have GNU tools and we do use Linux ... because that is what
> Arch is based on.
Exactly.
> I have no problem having GNU-specific features in there, as we can rely
> on GNU find (or GNU whatever) being present. The only exception is
> initramfs - however, no shell script is used in both the system and
> initramfs
That is one reason I helped write and use dracut (the new initramfs
generation framework that Fedora 12 and 13 use, and that RHEL6 will
use) -- instead of using busybox/klobc/whatever, it trades a slightly
larger initramfs for the convienence of using exactly the same
binaries the installed system uses for everything. I will happily
trade a few hundred K larger initramfs for not having to maintain
another toolchain and userspace just for the initramfs.
