Oops! You caught me (or did you?).

You say you can't non-reversably encrypt a password in a cookie. I stand
corrected. The Ipswitch error/oversight/ignorance/laziness that's always
bothered me is their reversable encryption of the password on the server
itself.

Heck, the algorithm to decrypt the passwords is very publicly available!

The user passwords on the server really should be non-reversably encrypted
(I know all the passwords I save locally for websites I manage are
non-reversably encrypted - Backup tapes can be stolen more easily than
servers!).

But - (does this sound like a cop out?) - Why can't the non-reversably
encrypted password in it's encrypted form be stored in the cookie?

We'd have:

 a) You can use your "regular" password, and no one compromising the Imail
server, a backup tape, a sniffer trace (if not SSL) or your stash of cookies
could figure out what your password was.

 b) A compromise of your cookies would yield ONLY access to that one IMail
server. No other applications could be accessed. Yes, I think that would
fly.

So, if Imail had non-reversable encryption of the password on the server,
and copied that to the cookie, the password wouldn't be in the "public"
domain and damage from a compromise would be limited to one email account.

Yeah, that's the ticket. (but it presupposes Imail has a non-reversable
password algorithm, which it doesn't - Darn).

Dan Barker

-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] Behalf Of Sanford
Whiteman
Sent: Thursday, July 21, 2005 5:09 PM
To: Dan Barker
Subject: Re[2]: [IMail Forum] Need some help


> The  password  should  be  non-reversably encrypted in the cookie if
> it's  there  at  all.  But  since Ipswitch doesn't understand how to
> non-reversably  encrypt  a  password, I don't think you'll be seeing
> this  feature.

That's  bit  FUDdy. A cardinal rule of cryptography is that encryption
that   requires   interpretation   and/or   decryption   without  user
intervention   is   naturally   insecure.   Performing  non-reversible
encryption  on  a  password,  then  accepting  the  encrypted form for
authentication  without  user  intervention  (as must be the case when
using  cookie-based  password  persistence),  is  far  from impressive
security.  True,  it  will  prohibit  someone  with  root  access from
reversing  the  password  and  using  it to authenticate against other
servers  if  --  as  indeed is common -- the same password is used for
multiple apps. But, in essence, encrypting the password in a cookie is
largely marketing to paranoid end-users who don't think things through
technically.  This  doesn't  mean  it  isn't  a worthwhile feature, as
customer  relations  are  always  important. But on a technical level,
it's pretty meek.

As  you  suggest,  not  persisting  the password locally _at all_ is a
better   solution.   I   think  the  existing  checkboxes  are  pretty
self-explanatory in that regard.

--Sandy


------------------------------------
Sanford Whiteman, Chief Technologist
Broadleaf Systems, a division of
Cypress Integrated Systems, Inc.
e-mail: [EMAIL PROTECTED]

SpamAssassin plugs into Declude!
  http://www.imprimia.com/products/software/freeutils/SPAMC32/download/relea
se/

Defuse Dictionary Attacks: Turn Exchange or IMail mailboxes into IMail
Aliases!
  http://www.imprimia.com/products/software/freeutils/exchange2aliases/downl
oad/release/
  http://www.imprimia.com/products/software/freeutils/ldap2aliases/download/
release/


To Unsubscribe: http://www.ipswitch.com/support/mailing-lists.html
List Archive: http://www.mail-archive.com/imail_forum%40list.ipswitch.com/
Knowledge Base/FAQ: http://www.ipswitch.com/support/IMail/


To Unsubscribe: http://www.ipswitch.com/support/mailing-lists.html
List Archive: http://www.mail-archive.com/imail_forum%40list.ipswitch.com/
Knowledge Base/FAQ: http://www.ipswitch.com/support/IMail/

Reply via email to