On Feb 28 13:32, random user wrote: > The changes regarding the user SID = group SID case look generally good > to me. > Thanks for considering the idea. > > I do wonder if it is best that the Everyone privileges would "leak" into the > group permission mode/mask tho, either at the Posix or ACL levels. They > don't > seem to for user SID != group SID, for those the oddball cases where > Everyone > has privileges the owner or group does not get a DENY in their ACL.
They do:
$ icacls xxx
xxx VINSCHEN\corinna:(R,W)
VINSCHEN\vinschen:(Rc,S,RA)
Everyone:(RX,W)
Successfully processed 1 files; Failed processing 0 files
$ ls -l xxx
-rwxrwxrwx 1 corinna vinschen 0 Feb 28 23:04 xxx
> Perhaps more importantly than that topic: I've noticed what seems a new
> issue in
> the behavior of the user SID != group SID case:
>
> rm x
> touch x # umask = 0077, user SID != primary group SID
> chmod 750 x
> chmod -x x # gets chmod: x: new permissions are rw-r-x---, not rw-r-----
This is perfectly fine, see the same sequence on Linux:
$ uname -sr
Linux 3.18.6-200.fc21.x86_64
$ umask 77
$ touch xxx
$ ls -l xxx
-rw------- 1 corinna vinschen 0 Feb 28 23:17 xxx
$ chmod 750 xxx
$ ls -l xxx
-rwxr-x--- 1 corinna vinschen 0 Feb 28 23:17 xxx
$ chmod -x xxx
chmod: xxx: new permissions are rw-r-x---, not rw-r-----
Corinna
--
Corinna Vinschen Please, send mails regarding Cygwin to
Cygwin Maintainer cygwin AT cygwin DOT com
Red Hat
pgpsc4MPsbTbs.pgp
Description: PGP signature
