Hello Helge,
On Sun, 26 Apr 2020 at 13:12, Helge Kreutzmann <debian@xxxxxxxxxxxxx> wrote:
>
> Hallo Michael,
> On Mon, Apr 20, 2020 at 11:40:26AM +0200, Michael Kerrisk (man-pages) wrote:
> > On 4/19/20 8:48 AM, Helge Kreutzmann wrote:
> > > **
> > >
> > > If newpath exists, this is already the reason for the failure?
> > >
> > > "If I<newpath> exists but the operation fails for some reason, B<rename>() "
> > > "guarantees to leave an instance of I<newpath> in place."
> >
> > I do not understand this report. Please elaborate.
>
> Trying to 2nd guess what the original translator, who unfortunately is
> no longer active, might have thought:
>
> If you want to rename but the new path already exists, how can there
> be other reasons for failure at all?
Because renaming onto an existing 'newpath' is permitted! Earlier, the
man page says:
If newpath already exists, it will be atomically replaced, so that
there is no point at which another process attempting to access
newpath will find it missing.
I think the translator must not have understood this.
> If this still does not make sense, please close this report without
> changes, of course.
Please close.
Thanks,
Michael
