The Death of Unity

I've been "game development adjacent" for most of my career. By that, I mean I'm not a game developer, I'm mostly a web developer, but I've done a lot of game development-like things in my projects over the years, including a 4 year stint where I built a VR app for foreign language training.

My impression during the last 20 years of my career is that game development lacks a lot to be desired in the "professionalism" side of software development. At least, on the indie side. I obviously don't know anything about what large studios do. But large studios aren't using Unity, either.

In the non-game development world, almost all of the best tools for development are free and open source. There are several different packages to choose from for any task, and they all pretty much integrate together. They support well-defined, open standards. Their developers work hard to make the software be stable across minor versions. And the tools--and thus the pipeline of developers--between small and large projects, small and large teams, are largely the same (caveat some scale-managing orchestration tools that one can get by without perfectly fine without users ever noticing on the small scale).

Before you might scoff-in-JavaScript at that statement, just spend some time trying to do indie game development. I've yet to encounter even a major version upgrade in a web dev tool that was as painful as even some patch upgrades were with Unity. Alternatives are vastly different. There's very little code one can share between even two .NET-oriented game engines. Thus picking one of those alternatives effectively shuts you out of the larger job market.

I used Unity heavily for about 5 years, after already having extensive, other software development experience. I very quickly grew a reputation with my peers for being very effective. Most of that effectiveness was just me writing tools to fix all the technical project management deficiencies in running a Unity project. Developing in Unity felt like the 6 months that I'll never get back of my early developer life doing FoxPro development. Unity is like Crystal Reports. Unity is like Dreamweaver. All those things in enterprise dev were also shitty, take-it-or-leave-it siloed tools. Yet I only ever encountered them in large corporations being ran by pointy-haired managers. And one would certainly never start a personal project using one of them.

Assuming Unity Pro ($2000 per seat per year) - first 1000000 installs are free, next 1000000 priced as follows: 0.15 * 100000 + 0.075 * 400000 + 0.03 * 500000 = 60000$ + Unity Pro for each seat.

I think part of the equation is that in that space most of the hard problems have been solved for the past 50 years - so while computing at whole is young, some common expectations have started to crystallize and this is apparent in the expected quality of open source tools.

However, in general IMO the software engineering -quality of an open source offering is inverse proportional to it's innovation and novelty. The more niche the domain, the likelier it is the open source there is sub-par per expectations - and not to fault the open source enthusiasts there! It just seems high quality open source either needs some diamond-level craftmanship from few contributors, OR a massive, massive userbase with professional interest aligned with the project.

In open source, another way to view this, is that in popular domains there are hundreds of competing teams creating similarish solutions to similarish well understood problem-solution domains, and then word-of-mouth quickly drives community to contribute those perceived as "highest standard".

From another point of view.

You still need fairly esoteric skillset to contribute to game tech. There are lots of things that simply are not taught that well together, and everyone needs to do lots of hard work to enmesh these concepts in their head to a whole. The work is not magical or rocket-science-hard, just something esoteric that manages to combine stuff from super hard or poorly defined domains like numerical computing and art into what is basically a frivolity. The implied hard work means you need same competence basically to contribute to game tech meaningfully where you could be curing cancer, doing actual rocket science or something of actual impact. As a field, games have an abysmal reputation of abusing talent and naivety. Anyone in their right mind absolutely not driven to develop games will choose any other field. Hence it's no wonder game tech is kind of ... random often.