cvs commit: src/bin/ls ls.1 ls.c

Maxime Henrion mux at FreeBSD.org
Thu Nov 10 18:49:30 GMT 2005


Ceri Davies wrote:
> On Thu, Nov 10, 2005 at 12:02:32AM +0000, Maxime Henrion wrote:
> > mux         2005-11-10 00:02:32 UTC
> > 
> >   FreeBSD src repository
> > 
> >   Modified files:
> >     bin/ls               ls.1 ls.c 
> >   Log:
> >   Add a -I option to disable the automatic -A flag for the super-user.
> >   
> >   PR:             bin/86710
> >   Submitted by:   Marcus Alves Grando
> >   MFC after:      3 days
> 
> For what it's worth, I think mnag was right in the PR when he said that
> this (at least the !getuid() bit) should be changed:
> 
> > | +	/* Root is -A automatically. */
> > | +	if (!getuid() && !f_nolistdot)
> > | +		f_listdot = 1;

I don't mind the change from if (!getuid()) to if (getuid() == 0) since
it is more correct WRT style(9), but what was suggested in the PR by
Giorgos (not mnag@) was to change all the if (!f_foo) and if (f_foo) to
an "expanded" form.  Or at least this is what mnag@ did in his second
patch.  Anyways, I don't think this is a reasonable change; the f_foo
variables are really boolean and it's simpler if we treat them as is,
and moreover, if people really want to do it this way, it can and should
be done in a separate commit.

As for the getuid() condition, feel free to change it.

> I'm also struggling for the motivation behind this option, but it
> doesn't actually affect me in any way whatsoever so I can't complain.

It seems both weird and frustrating that this automatic -A option for root
can't be disabled.  I personally don't mind it, but two different people
requested this feature on IRC, plus a third one sent a patch to
implement it so I don't see why it shouldn't get committed.  I'm
actually rather annoyed to have to justify committing such a minor ~15
lines patch.

Cheers,
Maxime


More information about the cvs-src mailing list