From Quake Hacks to Fintech Standards

From Quake to Fintech: The Strange Power of Obscure Knowledge

Dragon 32 computer
Quake / Fintech / ISO 20022 mash up

In the 1990s, I was obsessed with 3D graphics. Back then, computers were slow, so you had to be clever with maths. John Carmack and the id Software crew were legends for this - not just because of DOOM, but also Quake, which dropped in 1996 and introduced the infamous fast inverse square root hack (0x5f3759df). https://en.wikipedia.org/wiki/Fast_inverse_square_root

This trick abused the bit layout of IEEE 754 32-bit floating point numbers to get a lightning-fast approximation of 1/sqrt(x). It was the kind of hack that made you think: someone really understood the guts of the machine.

Fast forward to 2013, and I'm working at a water-pipe monitoring company called i2o. My day job was software engineering, but I often chatted to the hardware guys. One day, an engineer was pulling his hair out: numbers were coming back scrambled from his kit.

I looked at the data, thought for a second, and said:
"The byte ordering of your 32-bit floating point numbers is wrong."

Problem solved. He was happy, I was happy, and I'd just got to dip into some obscure IEEE 754 trivia that most people never need - but it mattered. https://steve.hollasch.net/cgindex/coding/ieeefloat.html

Same Story, Different Domain

Roll on 2017. I was at Open Banking in London, sharing an office with Pay.UK. By this point, I had a bit of a reputation for being "the guy you go to when the usual playbook runs out."

One of the Pay.UK engineers came over with a problem:
"Craig, I need to send more information in this message, but I've hardly got any bytes to work with."

It didn't take me long to see it: Base62 encoding would do the trick. I showed him, wrote up a two-page note, and suddenly his impossible problem wasn't impossible anymore.

What's the Point?

The link between Quake, IEEE 754, water pipes, and ISO 20022 is this: obscure engineering knowledge never really dies.

Different domains, same underlying lesson: when you understand the foundations, you can solve problems others can't even see.

That's why I love what I do. And it's why people bring me problems outside my "official" remit - because they know I'll figure it out.


Like this post?

I write about Open Banking, fintech, and the weird ways deep engineering knowledge keeps paying off.
Check out more Splig articles here. Or if your team has a "weird" problem, get in touch.

👋 Enjoyed the article?

Book a Call with Us