Default ACL: Semantics (again)

[James Buster]

On Oct 7, 12:17pm, "Peter J. Holzer" wrote:
} On 1999-10-06 15:33:22 -0700, James Buster wrote:
} > On Oct 6, 12:36pm, "Peter J. Holzer" wrote:
} > } I don't think so. 5.3.1.2 (2) says:
} > } 
} > }     Both the ACL_USER_OBJ ACL entry permission bits and the file owner
} > }     class permission bits of the access ACL are set to the intersection
} > }     of the default ACL's ACL_USER_OBJ permission bits and the file owner
} > }     class permission bits in mode. 
} > } 
} > } I couldn't find a definition of "file owner class permission bits of the
} > } access ACL", but from context I guess that it means "permission bits
} > } of all ACL_USER entries which match the file owner". So I think that
} > } user:andy: should be set to ---, too.
} > 
} > The "file owner class permission bits" correspond to the permissions
} > associated with the ACL_USER_OBJ entry.
} 
} Then why does it say "*Both* the ACL_USER_OBJ ACL entry permission bits
} and the file owner class permission bits"? This implies to me that these
} are two different things. But then English isn't my native language.

It says that because when an ACL is created, the Posix permission bits
must also be (logically) set. The phrase "file owner class permission bits"
refers to the Posix permission bits, not the ACL. That's why I'm sure
the words "of the access ACL" are incorrect, and somehow they got added
between drafts 15 and 17. When a file inherits a default ACL it doesn't
yet have an access ACL, so referring to it is wrong. The passage only
makes sense if those words are removed.

-- 
Planet Bog -- pools of toxic chemicals bubble under a choking
atomsphere of poisonous gases... but aside from that, it's not
much like Earth.
To Unsubscribe: send mail to majordomo at cyrus.watson.org
with "unsubscribe posix1e" in the body of the message