It’s now been over 10 years since I started getting a reputation for bitching and moaning about the state of Level Design Tools, starting with my first talk-into-mic youtube video, Level Editing In UE4 Needs To Catch Up To Quake 1, and then a bit later the wider-ranging sequel Ruminations Upon The Insidious Ubiquity Of Modular Mesh Based Level Design, both of which were, as I say in the second one, “well-received except where they were received horribly”. I’m relieved not to find these particularly cringey now if ever I watch them – I was a baby but I was right.
At the time, responses were split between confused anger and “finally somebody says it” exhausted relief. Over the years the former type have dissipated and the latter keep on rolling in; I rarely feel anymore when this comes up that I need to push these points. Maybe these worked and my point got across. I know they kicked off a lot of arguments.
It sounds improbable now, but at the time these were controversial in a way that really pissed a lot of people off. I’ll occasionally get served by Google Photos or Facebook a “memory” of a screenshot I took of someone getting real weird about it. Some of it I felt like I understood, some of it was just bewildering, and often it was condescending. I guess I was fighting in some of these people the idea that what is normal now and wasn’t before must be progress, and needs defending.
Some folks will have thought I was attacking them or their skillset, out of insecurity or even just confusion of terms – there’s an issue of “level design” meaning different things to different people, or being seen as in direct competition with other disciplines (like env art) when that’s only the case by avoidable pipeline misfortune. An Epic engineer or two were annoyed by the Quake 1 comparison, and I thought, why? That’s the last time you did any serious work on these. Do you realise you haven’t touched these tools in almost 20 years?
A real gem is the (iirc) Epic level designer for whom “the reality” was that he needed his tools to be bad and slow so that he’d be allocated more time to work on a given thing. This is the evil-aligned status-quo-defense version of an argument I’ve made before – if you can work twice as fast, the boss wants to give you twice the work, when they should let you do better work. The dude had fully internalised a busted situation as normal and fine.
Over the years I got varying levels of interest from Epic itself; assurance that it was being taken seriously and there were Big Things Coming in 2014, then nothing for years, a few aborted runs at a Modeling Mode that usually seemed to me to be predicated on misunderstandings. Whoever was the lead of the Modeling Team (or something) set up a meeting with me once to discuss the whole thing, which felt good and potentially fruitful, and when CubeGrid showed up later it felt like there was a throughline there.
I later had some pretty long twitter arguments with Ryan Schmidt, who was the main Modeling Mode person at Epic until he left in 2024, some of which felt productive and others needlessly aggro. He was doing important foundational work on modeling infrastructure, but was frustrated with the demands of (external) level designers. Reading between the lines, I think the reason he couldn’t present a good argument was that he had his orders, couldn’t say what they were, and the driving force behind them was not tools for game developers but the many concerns of UEFN, a concept which is now collapsing on contact with reality. He would so often take the tack of trying to tell level designers they were wrong to want things, though, that we all wasted a lot of time – I made the video below to illustrate some basics and he basically laughed incredulously at it. I still think he is a basically cool guy, though, who was under a lot of stress, and without him and his work in the engine, Scythe could not exist.
Meanwhile, at one point Unity reached out and paid me to spend time with several of the major Unity level design tooling products at the time, write up a big document explaining what level design needed and was missing, and make a recommendation as to which (if any) of these products they should acquire. I made a recommendation and advised against Probuilder; they acquired Probuilder.
When Source 2 became available (first via something called “Destinations”, then Dota2, then Half-Life Alyx) it became plain that Valve, unlike everybody else, had never dropped the ball on incredible level tools, and I did a massive twitter thread that became a blog post about it, which I know was handed around at Epic and Unity. Maybe it kicked off some discussions, but nothing came of it besides maybe making some of the drums relentlessly banged on by me and people like me seem less crackpot – nothing legitimises a theory in the eyes of the unthoughtful like Valve dropping a banger predicated on it.
Unfortunately, Valve having great tools doesn’t directly help anyone besides Valve, and hopes of the Source 2-based Garry’s Mod successor “s&box” being anyone’s salvation have been dashed by their being both technically dogshit and actively fascist. Similarly, there’s something called Anvil for Blender which looks good until you realise it’s by a vibe coder.
I genuinely think it would be an incredible idea to make Source 2 available. I suspect the main reason not to is they don’t want to take on a support burden (the reason given to me years ago about licensing their physics engine), but I would submit that you simply don’t have to. Make it apparent up front that the product is provided as-is, do your best, and if your best sucks, well, that’s Valve, man. It never stopped people making Source mods.
As for how level design tools are on Unreal now, we have Scythe! I’ve been testing it since before it was out, and it’s great, and it’s by a good dude. It’s essentially Hammer 2 inside of Unreal 5, and now all we have to worry about is Tim Sweeney steering his entire enterprise into the fucking sun. Everything I’ve heard is that Epic has its share of cult members, but is also full of people who absolutely hate the direction he’s taking, both with “metaverse” bullshit and “AI” bullshit, and everyone’s too scared to say anything. After all, the pay there is supposed to be insane; if you get fired from there it’s a permanent loss of income. It’s a shame what they’ve become.
Anyway, it’ll be interesting to see how this post ages. I don’t know if it’s ever worthwhile to make a bunch of people angry on the internet with a video saying something true about game development, certainly it doesn’t feel worthwhile, but this stuff is definitely easier to talk about now without getting your head bitten off, and maybe I was part of why. I mainly made this post so I had an easily-linked collection of these videos, but I guess it turned into a retrospective. I was a squeaky wheel for a long time, but now somebody has seen fit to grease me. Maybe I can be done talking about this now.
Leave a Reply