Re: XDG_CONFIG_HOME support in mbsync

[Andreas G. <an...@gr...>]

On Mon, Nov 30, 2020 at 07:56:36PM +0100, Oswald Buddenhagen wrote:
> On Mon, Nov 30, 2020 at 07:31:20PM +0100, Andreas Grapentin wrote:
> > On Sun, Nov 29, 2020 at 07:30:44PM +0100, Oswald Buddenhagen wrote:
> > > On Sun, Nov 29, 2020 at 05:37:43PM +0100, Andreas Grapentin wrote:
> > > > +			errno = 0;
> > > > that's unnecessary - the next error would overwrite it anyway, and
> > > no sane
> > > code would rely on this being zero without resetting it immediately before.
> > 
> > I don't like letting the state of errno leak to later code, if the
> > reported error has been handled. this can lead to unexpected output in
> > error reporting code in unrelated issues that don't overwrite errno,
> > such as a later call to sys_error.
> > 
> the thing is that this is a fool's errand. errno(3) on linux is rather
> explicit about what you should expect.

I think there's some minor benefit to this assignment, but if you want
it gone, it can go.

Updated patch is attached.

-A


-- 

------------------------------------------------------------------------------
my GPG Public Key:                 https://files.grapentin.org/.gpg/public.key
------------------------------------------------------------------------------