Re: mandb wrapper scripts
On Tue, Feb 08, 2000 at 08:43:01PM +0100, Ruud de Rooij wrote:
> Marcus Brinkmann <Marcus.Brinkmann@ruhr-uni-bochum.de> writes:
> > On Tue, Feb 08, 2000 at 04:50:12PM +0200, Fabrizio Polacco wrote:
> > > meanwhile, someone who tests this on hurd?
> > As Hurd su doesn't even have -c there is nothing to test here.
> What is the reason that the Linux su can't be used on the hurd? Or
> has it simply been not ported yet?
Well, I talked to Roland about it and he said we will probably punt our
minimal su and go for the su included in GNU shellutils, which supports -c
(IIRC). Anyway, the reasons why not to use linux su don't really matter
su is not standardized by POSIX or any other unixish standard (please
correct me if you know otherwise), and su implementations differ
significantly among unices. The only option that is fairly portable is "-c",
and I have no objection to use it freely.
So using non-standard options seems to be suboptimal, even if it works
on all Linux systems and can be made to work on all Debian systems (because
we have the source). I probably have to check out the complete discussion
now, to suggest other solutions to this problem :)
 The Hurd differs significantly in several low level aspects, and the
question: "why not use the XXX linux" is often answered by "because we can
do better", even if this sounds arrogant. For su, it is not that important
because we have native tools to manage authentification (setauth, rmauth,
addauth), (and our current su is just a wrapper around setauth --save).
The real question seems to be why Debian doesn't use the GNU su in shellutils.
`Rhubarb is no Egyptian god.' Debian http://www.debian.org Check Key server
Marcus Brinkmann GNU http://www.gnu.org for public PGP Key
Marcus.Brinkmann@ruhr-uni-bochum.de, email@example.com PGP Key ID 36E7CD09