Actually, understanding the vulnerabilities of SQL and other languages that access databases via websites are probably more important than assembly. (Unless you want to create the next Stuxnet for some unnamed agency.) The Army is funding significant work on making drone communication protocols unhackable. (You don’t need to be in the military to get DARPA grants, though.) Industry is doing similar work for automobile control systems, but many are behind the curve on thet.
I agree that strong CS with plenty of math is important for your BS degree. (Or just math focused on large primes.) Probably the main risk of not going back for a masters after being in industry is getting used to a lifestyle based on making real money, not so much forgetting the theory.
@Ynotgo I agree with you 100% on this. What is your opinion on the Computer Science program at the University of South Carolina? I am considering this strongly, in conjunction with a math major there.
@Ynotgo has a good assessment in post 60. However, that’s kind of what I was getting at with the rabbit and fence analogy. If you can force the application to chunk, what kinds of access does that maybe open up at the machine level - even for a couple milliseconds? Can that be exploited, even 1 out of 20 times? Knowing how the underlying machine behaves helps keep the expectations about magic protocols more realistic. This admittedly plays more into corporate than military mind sets.
It really is a systems problem, I think, meaning that all knowledge helps.