Thread by @lzsthw

Every time I read something by a cryptographer claiming to know when people are qualified to attempt cryptography software I go and see if they said *anything* about OpenSSL during the 20+ years it was obviously flawed:

soatok.blog/2021/02/09/cra…

Nope. Not a word.

In general nearly everyone who is currently working in cryptography accepted OpenSSL as correctly implemented for decades even though a basic analysis of the C code showed obvious flaws. They not only accepted OpenSSL but repeatedly yelled at people who didn't use it.

If you wanted to study crypto and implemented existing proven algorithms out came the Cryptography Experts yelling at you to "just use OpenSSL". It was *the* only implementation allowed. Then this happened:

heartbleed.com

Whoops! I guess everyone was wrong!

Not only did the cryptography experts suddenly become very quiet about how they were seemingly unqualified to evaluate the most basic flaws in C code, they *continued* to claim heartbleed was the *only* flaw. Nope:

openbsd.org/papers/bsdcan1…

To this day you'll have supposedly qualified cryptographers who still won't admit that OpenSSL was flawed, and everyone pushed it on the planet unjustifiably.

Then that same expert will yell like a rabid dog at anyone "rolling their own crypto" claiming they're saving the world.

So I'm adopting a new rule. Any cryptographer who either yells at someone for writing crypto code or rubber stamps their buddy's crypto code has to post a full blog post explaining why they missed the obvious flaws in OpenSSL and sign it with a formal apology.