>>>>> "Ethan" == Ethan Benson <[EMAIL PROTECTED]> writes:
Ethan> your test mail never entered the mail queue, since you
Ethan> don't have a world writable maildrop (this is not
Ethan> /var/mail) it was impossible for the message to get
Ethan> inserted into the postfix mail queue.
Oops, in that case I was obviously confused. I thought maildrop
was the last stage, not the first. Oh, I see:
/usr/sbin/maildrop != /usr/bin/maildrop
when I did "man maildrop" I got the second one. Oh well... I guess it
just proves, that no, my computer cannot read my mind ;-).
Ethan> no, this is a security feature of the kernel, it is
Ethan> actually rather common on unix like systems and i think its
Ethan> defined by either posix or SUS but i could be mistaken. s
If I had access to chgrp, then I also have access to chmod, so I don't
see any security benifit.
Ethan> bits are also removed if the file is written to (so a world
Ethan> writable setuid program can't be turned into a world
Ethan> writable setuid /bin/sh by simply doing a cat /bin/sh >
Ethan> /usr/bin/setuid)
false sense of security? Again, if you can write to the file, chances
are you can reset the s bit, too.
Ethan> in the future if you have to change this quickly do
Ethan> something like:
Ethan> chgrp postdrop /usr/bin/postdrop ; chmod g+s
Ethan> /usr/bin/postdrop
But then you can't do
find / -gid 104 | xargs chgrp 105
so the way I see it, this "security measure" actually is a "security
risk" because you have to replace the simple command line above with
something more complicated that saves/restores the permissions.
--
Brian May <[EMAIL PROTECTED]>
