Secure coding education

[James Walden <jwalden () eecs utoledo edu>]

"And in the one area in which their recommendations could make a long term 
difference -- upgrading computer science courses so no one graduates if they 
have not had secure programming skills and knowledge inculcated in them, the 
document provides no effective mechanism," Paller continued. "It's terrible 
when the industry says 'wait for us, we'll solve the problem,' and then 
delivers no effective proposals."



I agree that the computer science curriculum needs to add a focus on 
software security. If you look at the ACM/IEEE Computing Curricula 2001, 
you'll find security requirements under the areas of networking and 
operating systems, but not under the areas of programming fundamentals 
or software engineering.  I attempt to integrate software security into 
all of my courses, teaching secure coding techniques that are relevant 
to the subject being taught (input validation everywhere, buffer 
overflows when using C/C++, race conditions when discussing multitasking 
OSes, etc.) and secure design techniques in software engineering, as 
well as teaching a separate computer security elective.


If someone could send me a copy of the report on computer security 
education, I'd appreciate it.  I'm curious to see what their plan is, so 
I could see why it's not considered effective.  The largest issue I can 
see without having read the report is the problem of adding any new 
subject matter to the curricula.  CS students already have an extremely 
heavy load of required courses, and so it will likely be necessary to 
remove other material in order to add software security. 

Software security is also not an easy subject for students to 
understand, as it often requires integrating knowledge from multiple 
areas--software design, programming languages and libraries, operating 
systems, and computer architecture--in order to understand a particular 
attack, such as the common buffer overflow.   Teaching students to 
exploit software vulnerabilities is an essential tool for aiding student 
understanding, as constructing an exploit requires them integrate these 
disparate areas of knowledge.


--
James Walden, Ph.D.
Visiting Assistant Professor of EECS
The University of Toledo @ LCCC
http://www.eecs.utoledo.edu/~jwalden/
[EMAIL PROTECTED]

[Ed. A thousand pardons for taking so long to get this posting through
the queue.  I had an ill-timed disk failure on Sunday, just before leaving
home for San Francisco and wasn't able to get back to the local queue
until this morning.  Gotta love a good backup system...  KRvW]