After watching the yearly stream on UE I saw how old UEs used to work and that systems would eventually be replaced with improved versions (like Niagara, Sequence etc). However, there is always a tendency to go with “its weird, but it has better computational cost” (clear example the cube you had to curve out of instead of an empty area to build), and even today, this approach has translated to UE4, UE5 and now UEFN designs over the editor’s UI and naming. Now we are trying to learn a new workflow (which at this point the editor failed already, because if you need to learn something it means you couldn’t find it yourself first) then we see an obscure setting or process that just deviates from the goal and usually adds more problems to the mix. I am sure it allows games to run greatly, but you should start asking “is this toxic for the user”? Because if toxic accumulates, then bugs do too (most bugs occur from constant stops of work, better flow reduces them) and then you end up with a buggy mess that runs great.
A couple of examples of convolution are spinning an actor 360 degrees on the floor, in a sequence (in UEFN, and for extra fun, make it a decal) or changing the sort order of two emitters inside Niagara. Oh, and if you want to have custom architecture support build mechanics you are out of luck, because half of the settings in blueprint props don’t do what you think they do. This is not a problem in functionality though, it’s a problem of communication and I ll elaborate more below.
So today UE is the most powerful mainstream game engine in the market, yet it lacks so much on how user-friendly it is. If an engine is designed for mainstream use (and now UGC which requires an even greater level of ease and intuitive UI), how come there hasn’t been QOL improvements in basic things like translation in editor, animation in sequencer, or the import and creation of assets? Why when I move the gizmo off screen for an object I have selected, I can’t select it? Why do I need to select an actor and then manually select all of its attachments so I can select them together? Why can’t I select an actor inside a group? Why can’t I type whatever I want into a filename on the engine files? Why cant I paste where I am located in space, in the viewport? Why can’t I switch to isometric or orthographic views? Is it worth getting that computational reward from a “Space” character in an age with such computational prowess? Because it feels to me that this focus on performance has taken away from the fun of making games. If I spend 5 times the time I would spend simply importing a normal asset to optimize it, by the time I am done I would have missed 4 times creative fun I could have had. If you guys spent 1/3rd of the time you spend on optimizing performance to optimize the experience of the user when they build, the gamejam engine scale would tip massively on your side. Everytime I build something in Unity I am reminded of these issues again and again, so why not create an initiative where you say “We want to introduce features that will allow you to build smart, meaning fast and with the same result”. Add hotkeys, shortcuts, have the engine expect and resolve rather than warn and block. See discord, which has brilliant design that is made to not waste your time. This is what tools are for anyway, and they say that the tools make the craftsman.
Another thing is that since this engine is extremely powerful, all of that power goes unutilized because there is no proper and easy way to understand what half of the settings say/mean because their names are not organized in basic/advanced (you find a useful utilized basic setting after 4-5 advanced underutilized ones). And you have to browse clutter of unused choices everytime. Devices have partly solved that, and using the favorite function when it works is a great tool for speed (I have set most used settings in devices as favorites to see them often and change). You could even go a step further and add AI to understand and predict the settings you use the most, or change a setting in another window when you do X function, with a prompt. If you guys compare the clicks it takes to import a pixel style texture in UE/Unity and have it game ready, you ll understand what I mean (then scale that to 50000 to make a 2-year project and you ll see the difference in productivity).
Also why isn’t documentation embedded into tooltips that advice about the technicality of each setting, rather than being descriptive about its name? For example, draw calls, which workflow is the most optimal (blueprints with 1 mesh and swapping materials, BP with fixed meshes and materials, static meshes, dynamic actors, verse generated actors, etc.) Where can I find this info? Where can I know what a dynamic actor is? Where does it say what a blueprint does and why should I use it? I don’t want to browse a 40 paragraph page, I want a simple paragraph that is condensed and knows what I am looking for, directly in the editor itself. Dont waste our time, we need to help make games.
Anyway back to the point, the process of doing something in UEFN should be like Creative. A goal, a process, a result. Where does it get lost? Now lets say you want a cube to move. In creative you place a prop mover, and object moves. In UEFN, you have to place a cinematic sequence device, link it to a level sequence, link the cube through the sequence, set it to autoplay and you have to manually set the keyframes of movement. But then, if your world is streamed, you have to also set the cube to not stream or it wont show the same for two players. Now if you see both approaches, one was straightforward and related to the goal in hand, the other was lengthy, deviated from the original process, added friction and requires prior knowledge to do correctly.
Gimmicks should either be communicated or be avoided. If world streaming desyncs all sequences, that’s a known issue or if intended, then a gimmick of the engine. If Landscape doesn’t allow for XYZ sculpting, that’s also a gimmick of the engine that is not communicated, because the moment you are trying to sculpt an inverted island, everything is fighting you and you don’t know why.
What I am getting it at is that the engine has to steer clear of gimmicks and clearly communicate how to do things. The user doesn’t care how the engine works, as long as the process doesn’t create problems for the engine which then pass to the user. And I am full of these problems now, trying to make a simple game while optimizing my layout as if I am making the next AAA title. As I mentioned in my previous post on making UEFN (and in extension UE) a supertool, the functionality of the editor has to add clarity and through user testing spot the difficulties and learn curves and aid the user when need be through QOL AI/tools/features or at least some message/tooltip. If UEFN throws corruption messages when all I do is sync the project from URC from another team member just because they named a folder something that verse doesn’t like in asset replication, this is just a chain of events that creates chaos and wastes everyone’s time.
Simplicity is key. If you need more than one sentence to explain something to a user, you need to work on that tool more, until it’s simple enough to do so.
Similar to how we treat players like babies, you have to treat UGC devs like high-schoolers who know nothing of the prior history of the engine but have the passion to stick around and build if they feel included.
Anyway, I ll stop it here, for now, this should bring the point across. UI/UX needs to step up their game and revert 20 years of unorthodox fixation on performance (be extra awesome than what they already are), cause they are carrying us all on their backs and we depend from their tools so much for our games. If you need any clarification of the above, hit me up, otherwise until next time! beep
2 posts - 2 participants