UEFN lives between Fortnite Creative and UE5. However, the needs of the developers of UEFN I would argue are more aligned with the users of UE5 than players of Fortnite. In that vein, I believe UEFN should have a proper version testing and release cycle.
Failures of the current model:
- Communication: With the power of two image issue the only option is go or no go. Without betas, slow rollouts, implicit feedback cycles it creates a lot of pain for both the devs and the Epic side. For changes, Epic can only say ‘this is about to happen’ and devs can only say ‘no, I don’t want things to break’.
- Bugs: The Save point device and Accolade device both had major issues. For Accolade device this was enough for it being disabled for a period. This will continue to occur if all testing prior to release is only internal to Epic. This will only get worse are projects built on UEFN get more complex and the tools provided get more comprehensive. My personal frustration is the sphere trace change (Control Rig SphereTrace Function is Now Disallowed) that completely broke my project and there has been no Epic recognition of the issue.
With 18 months since the 2.0 release, there are now projects with hundreds of hours or more invested in them. I believe there is more an obligation on the part of Epic to support the developers here. We cannot just use an older version like one can with UE. We are forced to work in lockstep with each release and there needs to be investment to support us in doing that.
A solution:
I think it is pretty simple (in concept) with two requirements:
- Test Upcoming Version The ability for devs to develop against a release before it is pushed to Fortnite, with a time window for Epic fixing dev discovered issues
- Stability Commitment Some policy around stability, with incremental builds focusing on fixes and large numbers on new features
I do recognize the challenges in building out a solution here, and how it is a whole new capability to have a ‘beta’ fortnite. But, I think it should be clear that the current pathway is a bit untenable and will only get worse. At the point that people are building dramatically new game types or when a UEFN project captures a meaningful percent of the user base it is too late. With new game types it becomes too hard for Epic to internally find bugs. With a popular UEFN project the cost of breaking it via a release update becomes too large.
Now is the time to invest in a solution here.