History of Ironclad, a Common Lisp Crypto Library

It is refreshing to see the author of crypto code so honest about limitations of the code.

Appreciated the article. I've done something sort of similar with a Lisp dialect, Chicken Scheme, partly as a learning tool. Implementing SHA1 turned out not to be too difficult, though doing something more general is a whole other matter.

Working on implementing AES-CTR came really close to success, but there's some error in it. One of these days I'll have to get back to it and finish it. Of course, it is difficult to get it right.

I've had the idea that if it is correct in Scheme, and as Chicken does, then compiled to very regular C, the ultimate native-code program is much less likely to have the errors of hand-coded C as we've seen in OpenSSL, GnuTLS and others. Probably not a brand new idea, but certainly haven't seen it "in the wild".