RE: [MiNT] drive letters (was: Kernal questions)

["Julian Reschke" <reschke@medicaldataservice.de>]

> From: owner-mint@fishpool.com [mailto:owner-mint@fishpool.com]On Behalf
> Of Martin-Eric Racine
> Sent: Monday, March 08, 1999 9:24 AM
> To: Julian Reschke
> Cc: Frank Naumann; MiNT List
> Subject: RE: [MiNT] drive letters (was: Kernal questions)
>
>
> > > > > I would propose to just allow Fattrib() on the u: drive. This way,
> > > > > interested people can use chmod.ttp during system startup
> to hide the
> > > > > entries if they want to. Should be a painless change.
> > >
> > > As you know, the purpose of chmod is to restrict access (or
> hide) certain
> > > directories.  In the case of / it belongs to root, so if we
> allow a chmod
> > > on U: we could easily end up locking access to the whole filesystem.
> >
> > I don't see how this would be a problem if we support only the
> "hidden" flag
> > for Fattrib(), and if this ability remains restricted to the super user.
>
> If root hides / from itself, then who can still access it?  Nobody.

Everybody. "Hiding" is just setting a flag. The entry remains accessible to
all GEMDOS calls.

> Same problem if / is owned by root and root does "chmod 100 /" or
> something similarly stupid.  Then the whole filesystem is gone.....

Nobody has suggested that. It was suggested that Fattrib() supports the
"hidden" flag for entries in the root of U:\. That's it.

> Quite frankly, it would be much better to simply remove the parts
> that mount the logical drives into / from the kernel sources.

Frankly, this will break existing software for purely cosmetic purposes and
thus is inacceptable.