On 6/10/05, Rodrigo Barbosa <rodrigob@xxxxxxxxxxxxxxx> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> On Fri, Jun 10, 2005 at 09:01:31PM +0200, Maciej ?enczykowski wrote:
> >
> > title="$(sed 's/ release.*$//' < /etc/redhat-release | head -n 1)
> > ($version)"
>
> There is no need to use "head", since sed can handle that on its own.
>
> Just make it 's/ release.*$//;q'. It will only process 1 line from
> the input file, even if 's' doesn't match anything.
>
These both seem flawed and could be improved by doing the search for
CentOS to handle the case where someone puts a new line at the
beginning of the file instead of the end. Unless of course the
various software that require that entry in the /etc/redhat-release
require it to be the last line in the file - I wouldn't know, I don't
use them.
Of course, that would be yet another deviation from the upstream SRPM
which only helps a few corner cases...seems not worth it.
Greg
