It’s been a while since I made one of those, but I will be short since most of the things I want to say are applicable from the previous posts I made about UEFN as a whole (look up my profile >interactions > filter posts on feedback to see them)
Quick list prioritized, these are all things UEFN needs to address to “start” being a competitive platform for development and games:
- Stability: This again is the top of the list, NOTHING should break on an update. The test would be to leave a complex game for 6 months and no regression that is higher than medium priority would appear between builds. Since UEFN forces the “always on live” flow, Epic needs to adapt and innovate on making absolutely sure nothing breaks on an update, even if it seems impossible. Here are some initiatives to help:
- Creators have a platform to communicate disruption to the player with clear signage and verbiage. Things like little icons that we can add on our discovery thumbnails with a click of a button that says “Under maintenance” and when users hover it explains the game is under maintenance for an update transition. Also creators can add a note on their description about such disruptions and mention an ETA on a fix.
- AI to be used to vet builds and spot any differences between them. Have them run the builds and check for differences on around 100 maps from the ecosystem. Also, have them perform all editor actions possible, including live edit and creative, on both a fresh install AND a machine that updates (as some issues appear that way too) and also check for differences there from the previous build. If done thoroughly and the human user understands how to flag the differences and convert them into tickets, this will probably catch 70% of the regressions that happen on an update (as this figure is easy to spot on basic functionality, 30% only happens on scaled projects). Same with gameplay feel, lighting, devices behaving as intended, URC, auditing/mem calc, HLODs, spatial loading, placing all props and prefabs and checking their LODs and look, loading times for everything in the editor and game, matchmaking, etc.
-
Discover: I ll keep it simple since things have already been said about Discovery. The test is simple: with California wages a 4-6 team with a game that sits between a niche and mainstream and a development time of 3 months should make ROI on release. This would be ANY game of that scenario, except if the game is severely broken or unfun for its niche. Tweaking discovery so it can enable that, or adding a different system for engagement that doesnt bleed players dry or disrupts their experience, would be great. And the ROI should scale based on how much quality and hard work has gone into the game, don’t make us doubt ourselves and our ability. We are not here for impressions, but creating sentiments. This is not a social network and everyone in Epic should be told this again and again. Oh, and steer away from FOMO, it creates cases where the games don’t deliver and the player doesnt have a choice.
-
Fun ecosystem: Simple stuff, the test here is that a player WANTS to play at least 3-5 experiences a day, without thinking about Battle pass exp/cosmetic reward or what their friends are playing. They just want to play to have fun, experience something and feel. To do this, embrace variety, remove tie ins between discovery and game details (except to create excitement), and embrace games with character.
-
Fun to build: If a dev is not having fun making a game, they wont make a fun game. The test here has two parts: If a new user can make a game in a day and have fun (without guidance), thats a pass on test #1. Test #2 is to have an experienced user not be distracted for 30 mins of working with the editor with ANY technical problem that is not part of the game but part of the engine and tools. Literally fix everything there so there is no stalling, no weird thing you have to do to do XYZ, validation warns proactively not reactively, no bug preventing you from completing a task, no inverted or missing controls, no loading screens on plugins we don’t use, and most of all, no bloat. Optimise the tools to be fast and efficient. This will create the fun for development, as the timing between the idea and the product will shorten. Small projects should be made in 1-2 weeks, medium a month, large projects 3 months. Currently its 3 times that due to bloat, instability, and unnecessary dev work. Also let us use AI to perform repetitive tasks, like adding boxes as collision on a complex object that is not a rectangle, or retargeting characters, or animating them, or creating verse snippets to perform game logic, or set dressing using FN props in a defined zone. AI has so much potential to help us be faster. The utopian scenario is that one user without budget can make any game they want while having fun, express themselves, in a day. This will democratize game design and let everyone build their dream games without having to worry how. The goal is to build them, after all.
EDIT: Launching live edit with unsaved changes still stops the process a while later after you tell it to. This is the worst UX of the editor and it hasnt been addressed after constant flagging. When opening live edit, autosave if autosave is turned on in editor preferences and remove the prompt to save.
-
Expanding games. The last but certainly not least issue we always encounter is the finite end to every project. We are not able to expand our games to the scope of the project, but always to the limitations that exist. Let us connect levels together and pass variables between them, let us make campaigns, let us unload logic, give us proper auditing tools and memory calculation breakdowns with heatmaps or more detailed information. This part has actually been worse tooling wise since UEFN came out; even if we got more memory to work with, we dont know where that memory goes once we build (we used to know before the system was changed).
And lastly, this is a UGC ecosystem. Have it be run by people who have previous experience on UGC. You dont put a plumber to work on a rollercoaster just because both are made of metal. As UGC becomes a mature industry, showcase internally the people who have experience with such platforms (and def not social networks, as I ve said above) which I know you have because I have been keeping track of your workforce. Those people are key to the future, find them a spot where they have the power to act and keep them stable for at least 5 years so they can work their magic.
For UEFN creators, if you are reading this post, which point do you agree the most, vote below:
- Stability
- Discovery/ROI
- Fun for player
- Fun to develop (less friction)
- Expanding games
- Other (post in comment)
