From: Ted Ts'o <tytso@mit.edu>
To: wayne cripps <wbc@cs.dartmouth.edu>
Cc: linux-ext4@vger.kernel.org
Subject: Re: bsd groups
Date: Wed, 17 Nov 2010 20:41:54 -0500 [thread overview]
Message-ID: <20101118014154.GN3290@thunk.org> (raw)
In-Reply-To: <B289F406-D03C-499D-9F3B-EAF5B6E5DEF2@cs.dartmouth.edu>
On Tue, Oct 26, 2010 at 02:05:00PM -0400, wayne cripps wrote:
>
> Our department - Computer Science, at Dartmouth College, depends
> on bsd groups for our projects. Having bsd groups means one less
> thing for the undergrads and their TAs to have to keep track of.
> Since we have used them forever, people are used to that behavior as
> default.
So if people really insist on using it, then we'll keep it.
But one thing which I don't understand; why can't you just set the
setgid bit on the shared spaces? This causes newly created files to
have the same group id as the directory, and newly created
subdirectories to have the set gid bit. Set the group id at the top
directory of the project's directory hierarchy, and set the setgid
bit, and all newly created files will inherit the group id of the
directory --- and all newly created subdirectory will inherit the
group id as well as the setgid bit.
This is the System V scheme, which is much more flexible than the BSD
scheme, since you can set control whether you have the BSD behaviour
or the original System V unix behaviour, which is to always use the
primary group ID of the creator.
If you set the setgid bit on all directories (chmod -R g+s /mnt), then
you'll effectively have the same behavior as the grpid mount option.
What we may do is add a pointer to a web page with this explanation so
that people can understand there is a better alternative. Does this
work for you?
- Ted
prev parent reply other threads:[~2010-11-18 1:41 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-10-26 18:05 bsd groups wayne cripps
2010-11-18 1:41 ` Ted Ts'o [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20101118014154.GN3290@thunk.org \
--to=tytso@mit.edu \
--cc=linux-ext4@vger.kernel.org \
--cc=wbc@cs.dartmouth.edu \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).