8 August 1997
Source: Mail list cryptography@c2.net

See also the National Security Agency's MISSI site: http://www.nsa.gov:8080/programs/missi/

And the NSA's FORTEZZA site at: http://www.armadillo.huntsville.al.us/

To: cryptography@c2.net
Date: Thu, 7 Aug 1997 21:27:20 -0700 (PDT)
From: Phil Karn <karn@qualcomm.com>
Subject: Fortezza 2.0 Cancellation

------- Start of forwarded message -------

>From: "Fritsch, Kenneth M." <kmfrits@missi.ncsc.mil>
[lengthy recipient list deleted]
>Subject: Fortezza 2.0 Cancellation
>Date: Thu, 07 Aug 97 09:43:00 EDT
>Encoding: 73 TEXT
>     Thank you for your interest in the Fortezza 2.0 architecture! However, 
>at this time the Fortezza 2.0 Architecture development effort has been 
>cancelled by the National Security Agency, X2. Many factors contributed to 
>the cancellation of the effort. I wanted to explain the situation to our 
>industry partners for you to gain better insight of our current direction 
>and provide you the reasoning for this change in direction. The viability of 
>the architecture was not in question, only the Government's resources and 
>direction were issues. The predominant deciding factors were:
>1. Our new corporate direction for network security, the commercialization 
>strategy, was inconsistent with the direction taken in the Fortezza 2.0 
>architecture work. We will now try to drive compatibility at the CAPI 
>(Cryptographic Application Programmer's Interface) instead of the lower 
>levels, namely the libraries, drivers, and card interfaces. We desire 
>complete solutions where the developer of network products will select and 
>maintain those interfaces independent of the CAPI level. The Fortezza 2.0 
>architecture work was in the process of defining these lower levels.
>2. NSA can no longer afford to pay for the developments to drive the 
>marketplace. We do not have a large enough market to substantially drive the 
>commercial market. Also, with that in mind, we do not have the manpower or 
>budget to see the Fortezza 2.0 architecture through to completion in the 
>commercial market. Therefore, our resources have been redirected to other 
>3. With the use of commercial crypto for classified systems using a layered 
>security approach, much of the applicability for the Fortezza 2.0 
>architecture for the Type 1 market niche has been eliminated. We can not 
>sufficiently address changes in the commercial market and our specific 
>classified system needs are being addressed in these other non-traditional 
>4. Industry has already independently started to adopt pieces and principles 
>of the Fortezza 2.0 architecture in the multi-application Fortezza cards and 
>more specifically in the PC/SC (Personal Computer/Smart Card) standards. 
>This seems to be the direction technology is taking - without our lead.
>The National Security Agency will now focus our efforts on interoperability, 
>key management infrastructure support, and high level testing - to name a 
>few related areas that will be stressed. The Fortezza 2.0 architecture team, 
>however, performed valuable work which is still viable in today's commercial 
>marketplace. I can provide the architecture documents to allow synergy to 
>occur in the private sector with our industry partners. There were many good 
>ideas in the architecture that we wanted to share with any interested party 
>to help advance technology. The architecture had many desirable features to 
>   - Multiple non-cooperating applications
>   - Multiple Users
>   - Multiple Tokens (includes PCMCIA, Smartcard, Software)
>   - Extensible functionality (Ability to add new algorithms, new key 
>   - Fixed structured signalling
>   - Improved performance over existing Fortezza architecture
>   - Improved software architecture to allow independence of software 
>modules for
>     easier upgrades.
>   - Interoperability at the CAPI/CSP level.
>   The output of the Fortezza 2.0 architecture is contained primarily in two 
>documents, The ICD and the connection manager ICD. I will assume that you 
>have received the output of the FACT team previously (Communications 
>architecture, and software architecture). The ICD is about 85% complete and 
>contains all of the information of the architecture details and the 
>connection manager contains information that allows for sharing of resources 
>for multiple applications/users/tokens. If you have any additional 
>questions, please ask and I will do my best to provide answers. In the 
>interest of
>allowing this e-mail to go  out efficiently I have not attached the 
>documents. If you would like to get the documents - please ask. Thanks!
>Ken Fritsch
------- End of forwarded message -------