As shown in the roadmap for Unity 6 at Unite: https://youtu.be/pq3QokizOTQ?t=692 there’s a planned WASM stripping tool. That said, it was unclear if it was already available for Unity 6 or if its planned for Unity 6.1+.
Also hoping if it was going to be available for the Web platform outside of Facebook instant games (I don’t see why it shouldn’t be accessible for all web exports…)
It’s a great initiative. I’ve been asked to aggressively target lower build sizes for WebGL. The lowest we ever got to was 4MB, because at the time the splash screen took up 3MB of build space even if you had a Pro License and didn’t include it
Thanks for your reply! Like I posted in the Unity 6 announcement post; an ol’coleague of Unity (Aras!) has posted a very interesting writeup of the Web Platform build sizes. As you know, this is a major pain point for the platform - as it directly correlates to the time-to-play in games.
Thanks for calling this out! Some of the engineers on our team are actually active on that thread and looking into the suggestions. It’s also worth noting that a good amount of these recommendations have been captured by the updates to our Unity 6 Web Optimizations documents that can be found here:
I would suggest adding disabling Post-Processing as a suggestion formally as it is by far the biggest contributor to the build size of a URP project.
Additionally, there’s a few caveats that should be noted in the “Disable a built-in package” for big packages like Terrain that can’t be disabled due to dependencies. Or just a general comment that disabling packages with dependencies will not have any effect.
And finally, if Unity could report / test on minimal build-sizes that would be a great addon.
Ex. Minimal build size for Web 3D URP (with all packages stripped except bare minimum, such as UI and input = 3.2MB) for Unity 6000.1.
Even better would have it as a “template” for new projects. It gives us a target size reference & foundation to build-upon. And for you internally to track if new dependencies are added by mistake by other feature teams.