From: Adam Shostack <adam@homeport.org>
To: deviant@pooh-corner.com (The Deviant)
Message Hash: b69e4b3723143dd9c0669cb6672ee4f6530de7e568b2fb3c06c97f4105828010
Message ID: <199611171732.MAA05948@homeport.org>
Reply To: <Pine.LNX.3.94.961117172527.2314A-100000@random.sp.org>
UTC Datetime: 1996-11-17 17:37:09 UTC
Raw Date: Sun, 17 Nov 1996 09:37:09 -0800 (PST)
From: Adam Shostack <adam@homeport.org>
Date: Sun, 17 Nov 1996 09:37:09 -0800 (PST)
To: deviant@pooh-corner.com (The Deviant)
Subject: Re: RFC: A UNIX crypt(3) replacement
In-Reply-To: <Pine.LNX.3.94.961117172527.2314A-100000@random.sp.org>
Message-ID: <199611171732.MAA05948@homeport.org>
MIME-Version: 1.0
Content-Type: text/plain
Unless you're running yp, or if your wu-ftpd leaves a core
with the password entries still in memory, or sendmail can be used to
read any file on the system...
Belt *and* suspenders, and a lot more simplicity than wu-ftpd
or sendmail offers you.
Adam
The Deviant wrote:
| On Sun, 17 Nov 1996, Adam Shostack wrote:
| > The Deviant wrote:
| > | On Sat, 16 Nov 1996, Joshua E. Hill wrote:
| > | > I'm trying to think of a function to replace UNIX's crypt(3).
| > | > My design criteria are as follows:
| >
| > | Why? UNIX passwords with password shadowing are as secure as any password
| > | system is going to get. If your security holes are with passwords, its
| > | because your admin is to lazy to install needed security provissions, not
| > | because the system of checking passwords is bad.
| >
| > A longer salt would make running crack against a large
| > password file slower.
|
| While thats all well and good, it shouldn't be necisary. If passwords are
| shadowed, one must have root access before one can run crack against the
| password list, at which time it is innefective.
--
"It is seldom that liberty of any kind is lost all at once."
-Hume
Return to November 1996
Return to “The Deviant <deviant@pooh-corner.com>”