So, recently I finished reading the book Masters of Doom**** by David Kushner. It chronicles the rise of id software, the creators of the seminal video game Doom. Coincidentally, John Romero, one of said masters, has a new autobiography out now as well, titled, Doom Guy. I’ll have to pick that one up too!
As someone who’s worked in software for my entire career and who spent the better part of a decade and a half in the games industry, this book was hard to put down. I remember when these games were released and spent many nights playing deathmatch in Doom on a 2400 baud modem! There were a bunch of things that really struck me about this story that felt in stark contrast to the reality of software in the 2020s.
Subscribe
These guys shipped!
Modern games take years and years to build. Yes, they’re bigger. Yes, they have more content, but we also have way, way better tools than these guys did. They had to basically write everything from scratch. There was no Unreal engine or Unity. And yet, they shipped stuff multiple times per year. You would think that with modern toolchains and decades of learnings about the medium we could ship faster than we do.
They literally invented modern gaming.
Speaking of tools, it’s crazy to see the pace of innovation on display. We take genres like First Person Shooters and their conventions for granted now because, well, almost all AAA games for the past 25 years have been some variation on an FPS. John Carmack basically invented side-scrolling on the PC for Commander Keen, texture mapping for Wolfenstein, took that to the next level with Doom, and then essentially the modern 3D multiplayer engine for Quake.
No big deal.
How long could it possibly take, two weeks?
There’s a modern meme post where some person on Twitter asks a game developer how long it could possibly take to add multiplayer to a game, a couple weeks? And then anyone who’s ever worked on multiplayer (especially realtime multiplayer) falls over laughing because it’s so far out to lunch.
And yet…there’s a segment in Masters of Doom where they’ve finished the development of Doom, but realize they need to add multiplayer. And so Carmack does it in like…a few weeks. Bonkers.
Did I mention innovation?
Oh, and that multiplayer mode, Deathmatch (where you run around a map and try to kill as many of your opponents as possible), is basically the template for all the arena shooters, battle royale games (eg. Fortnite) and every major FPS (eg. CoD) to this day. It’s a mode that seems to just sort of happen during the development of Doom without a lot of planning (it is in fact literally an afterthought) and yet, it basically invented the entire modern AAA landscape from a design standpoint.
We won’t even get into the fact that they innovated on the distribution method by releasing these things as shareware, spearheaded modern level design, and spawned the LAN party.
It’s not all sunshine and lollipops.
The book outlines considerable strife between the two Johns as id rises in popularity and—spoiler alert—it all ultimately disintegrates in spectacular fashion. Also, the amount of “crunch” and deathmarch development involved in these iconic franchises would make modern dev teams cringe.
That culture of punishing 16-hour days, 7 days a week was present—even dominant—in game development through the 00s at least. It wasn’t until the 2010s that things started to improve, but crunch periods on large franchises still persist to this day.
The Good or Bad Old Days?
I’ll admit that reading this book made me a bit nostalgic for the days of shareware, BBSes, and writing software where you have to directly manipulate video memory to draw things on screen.
I was also nostalgic for the days of just doing things based on inspiration and instinct, that didn’t need to be A/B tested, or have market research or psychological studies done. You just had an itch you had to scratch and so you built it. That somehow seems more human. And in an age where AI threatens to automate and sterilize even more of our creative output, the human-ness of those old games like Doom, seems extra special.
The flip side of course is that modern tooling and access to information is so much better that it’s almost not even the same pursuit. In those days, you were scouring Dr. Dobb’s Journal for code listings you might be able to adapt. Without Stack Overflow, we’d spend weeks or months trying to figure out arcane things like how to draw sprites to the screen with transparency fast enough that it wouldn’t flicker.
And everyone who was working on these things was figuring it out for themselves. Almost nothing was shared. It was actually incredible that the source code to things like Wolfenstein was shared after a certain point, and those games were built with modding in mind. The book notes that Carmack is violently opposed to even the concept of software patents.
These days, I’m not sure many developers even think about how sprites are drawn to the screen. It just works. Cameras are just a standard, boring engine element—not an exercise in complex linear algebra trying to calculate how to project points from 3D space onto a 2D plane.
And that should be good. It should mean more mental space for creativity, innovation in the medium, and faster development times. But it doesn’t. Games today are absolutely more polished. More expansive. More accessible. But are they ultimately more fun? Has some unique human element been lost in the era where anyone can sit down and get an FPS running in mere minutes, yet somehow the innovation seems to have slowed?
I’m not sure. But I think there are still lessons to be learned from these now historic games, some valuable insights to be gained from reflecting on what we might’ve lost over time, and how we can bring that kind of energy and inspiration to modern development.
What do you think? Was there something more human about those old games? Let us know in the comments!
And if you liked this post, take a second to click the heart, share it with a friend, or leave a comment.
Subscribe
