This is the mail archive of the
cygwin-developers@cygwin.com
mailing list for the Cygwin project.
Re: ntea & chmod
On Aug 29 07:22, Pierre A. Humblet wrote:
> At 11:24 AM 8/29/2004 +0200, Corinna Vinschen wrote:
> >On Aug 28 23:38, Pierre A. Humblet wrote:
> >> pc.fs_has_ea () should be set properly as well in path.cc.
> >> Currently it's set to make sense for symlink (that's obsolete),
> >
> >Obsolete? Hmm, you're right, PC_CHECK_EA is never set anywhere, AFAICS.
> >Do we still need that code?
>
> No, get/set_symlink_ea can be retired, or at least #if'ed out
> and the tests about using them can be removed.
I agree.
> >I'm wondering though if it still makes sense to keep the
> >get/set_file_attribute functions as they are. Except for two cases,
> >the functions are either called for the ntsec or for the ntea case
> >exclusively. Perhaps it would make things clearer again, if we use
> >get/set_nt_attribute (resp. get_nt_object_attribute) and another
> >function pair get/set_ea_attribute.
>
> Yes, that's worth a try. We seem to duplicate the tests in the calling
> function anyway.
Yup, that's what I'm up to. We're testing everything twice. In a way
the test in get/set_file_attribute is useless because the calling
function has to make an educated choice, already.
> In the ntea case, shouldn't the attribute be set in ::open when a file
> is created, with the same kind of logic as in mkdir?
It's done in open_fs. Do you want to move it?
Corinna
--
Corinna Vinschen Please, send mails regarding Cygwin to
Cygwin Project Co-Leader mailto:cygwin@cygwin.com
Red Hat, Inc.