Is there a Generic "Doom Principle" of Programming Patterns?


As the software world is reeling over the log4j exploit (aka Log4Shell, aka CVE-2021-44228), I can’t help but feel a little tinge of unhealthy schadenfreude over the Java ecosystem. As I see it, Java’s infatuation with remote object directories and service providers directly led to this whack-a-mole-security-vulnerability reality that Java has been stuck in for decades.

It brings to mind Moxie Marlinspike’s 2011 article “The Cryptographic Doom Principle” where he laid out the following:

When it comes to designing secure protocols, I have a principle that goes like this: if you have to perform any cryptographic operation before verifying the MAC on a message you’ve received, it will somehow inevitably lead to doom.

He then detailed two examples of cryptographic exploits that demonstrated the rule before concluding:

even if these particular cases don’t apply, the general pattern will somehow inevitably cause trouble. It always does.

Not only was Moxie spot-on, I think the pattern of his analysis is generally applicable to most programming, not just cryptography.

Consider this generalization:

“Some patterns will somehow, sometime, inevitably lead to doom, always.”

There’s some subtly here. As I see it, the primary risk of such a Doomed Pattern is not that it will fail in every use case, but rather that the pattern will eventually and inevitably be at the root of one or more spectacular failures.

The Doom Principle sits at the nexus of “Code Smells” and “Tech Debt”. The reason we care about identifying “smelly code” is because we’re implicitly looking for a Doomed Pattern. The point of code smells is being able to identify code that could be problematic under the right circumstances, but isn’t necessarily bad, and could be fine in a particular case. In this way, a code smell is broader and a little more superficial. A Doomed Pattern on the other hand operates more at the level of “if there’s smoke there’s fire” or “rotting fish smell equals a rotting fish”.

I look at the Java ecosystem and interpret the world as follows:

Java: “We have shipped this awesome ecosystem of tools to make dynamic code discovery, loading, and execution across networks and organizations easy!”

Java, every day since: “We have identified some remote code execution vulnerabilities that need patched.”

After some more thinking, I would love to consider some other examples of Doomed Patterns. Future Tommy has a job to do.

comments powered by Disqus