Actually, I was so desperate for pen-pals who use digital IDs that I
signed my last posting.  Unfortunately, the WEDI listserve adds a text
trailer (telling you how to un-subscribe) at the bottom of the e-mail,
thus invalidating the signature.  Reliable S/MIME aware mail-clients
will report that the signature (not the certificate, though) is invalid
since the message has indeed been tampered with (by the list server).
Everything worked as it should!  I normally turn off signing when
posting to a listserve because either it (1) adds or modifies text, or
(2) forbids attachments (which signatures are implemented as).

Digital IDs will certainly expire - that's one of their "features."  But
you can certainly read archived e-mails even with an expired
certificate:  the e-mail client will simply warn you that the
certificate used for decrypting is expired before going ahead and
decrypting the message.  This has never been a problem: I simply save
the few certificates that I've ever owned (and thus I can decrypt all
messages that I've ever sent or received: my "sent" copies, like
messages sent to me, are encrypted using *my* public key).

Certainly, as I've acknowledged, there's a little tweaking here and
there to make this stuff work, but it's really no more complicated than
network settings and things like that. Centralized key escrow and
archive that will be appropriate in the corporate setting add even more
complexity.   But truthfully, I thought we were talking about
interoperability, not the admittedly extra (and predictable) expense of
key management.

I didn't quite follow the business about viruses, and wouldn't know why
using X.509 and S/MIME makes one more susceptible to them.

William J. Kammerer
Novannet, LLC.
+1 (614) 487-0320

----- Original Message -----
From: <[EMAIL PROTECTED]>
To: "William J. Kammerer" <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]>
Sent: Monday, 29 April, 2002 09:23 PM
Subject: RE: Transmitting Patient Information via Internet (Email)

Hate to tell you this, but my outlook reports your certificate invalid!
So much for theory. In practice, certificates have a significant number
of management and administration problems. They expire and can make
archived email inaccessible. They require significantly more help desk
and seat management time. Additionally, if you have a certificate, you
are vulnerable to its use by email based virus that send from your email
client once infected. At least web-based / internet browser viewed
secure email is secure and reliable based upon the server SSL
certificate (different than the distributed certificate that accompanies
regular email). Though again the problem with both is lack of Trust. The
network is as much the problem as anything. There are "sniffers" that
can capture email passing through most ISPs, and MIME can be broken. New
"harvester" worms infect your email client and copy/resend emails based
upon special criteria. In my opinion, only if there is a true trusted
network, can you reasonably assure compliance. Outlook is a great
product, but at this time, I personally do not see it as HIPAA compliant
for either the Exchange or Internet Email clients, where the email
travels over the web. Products like Lok, and any others like it offer a
fully secure and trusted solution.

-----Original Message-----
From: William J. Kammerer [mailto:[EMAIL PROTECTED]]
Sent: Monday, April 29, 2002 7:06 PM
To: [EMAIL PROTECTED]
Subject: Re: Transmitting Patient Information via Internet (Email)


I religiously sign using X.509 certificates (digital IDs), and have had
very few problems corresponding with either folks who have given me
their certificates (in which case e-mails can be encrypted), or those
who have no certificates (in which case, I merely sign). The latter
"un-certificated" folks will most likely receive my e-mail showing a red
ribbon indicating the message has been signed and giving them an
automatic means of importing my certificate (public key) - assuming they
are using an S/MIME compatible e-mail client.  Others (using AOL - which
doesn't support any standard e-mail protocols, let alone S/MIME - or
free web browser e-mail) will merely see a pkcs7-signature attachment,
which they can safely ignore.

My correspondents have used any number of e-mail clients:  Outlook,
Outlook Express or Netscape Communicator (on Windows or the Mac) and all
have worked flawlessly as far as signing and encryption are concerned.
Any number of encryption methods have been used among us, e.g., 40-bit
RC2, DES and Triple-DES, with nary a concern.

Sometimes, though, we have to futz a little to get a digital ID properly
associated with an address book entry.  But once that's done, secure
e-mail Digital IDs or certificates from any number of CAs - usually
Thawte or Verisign - have never caused a problem with interoperability
using these e-mail clients. I've even had correspondents (who don't
trust CAs for some reason) give me self-signed certificates, which I've
gotten to work easily.

The only serious problem I have run into is one zealous network
administrator at a correspondent's company who thinks pkcs7-signature
attachments are viruses, and has tuned the virus scanner to throw away
my signed missives:  I always have to remember to reset the "Sign"
button when sending to that company.  Other network administrators, I'm
sure, are annoyed my correspondents use encryption, as it gets in their
way of reading all incoming and outgoing e-mail in their copious free
time.

In short, any e-mail client which advertises itself as supporting S/MIME
has always seemed to work for me and my correspondents.  The few
technical gotchas are insignificant compared to the problem of getting
folks to try it out.  All the software they need for secure e-mail is
probably already sitting on their desktop:  it's just a social
engineering problem to deal with the resistance.

William J. Kammerer
Novannet, LLC.
+1 (614) 487-0320



**********************************************************************
To be removed from this list, go to: http://snip.wedi.org/unsubscribe.cfm?list=privacy
and enter your email address.

Reply via email to