As I believe I’ve made the decision to at least temporarily end my subscription to Unreal Engine 4 as of tonight - I went looking for threads where people might be talking about similar things, to gather my thoughts and maybe provide some feedback in the same place as others as to why I am doing so. I couldn’t really find much discussion to that effect; I found one old thread that was only passably relevant so I’ve instead decided to just make a new post.
I’ve been using Unreal Engine since UE2 (Still love UT2004!), and did professional development on UE3 and UDK as a designer. Career forces had me switch to another well-known engine, but with the events of GDC and the new pricing model I was VERY excited to return to Unreal Engine, feeling the spirit of a pricing model that would make possible what I truly wanted to accomplish as an indie developer.
Out the gate I was so impressed with the rendering performance and visual effects that there was no question I’d be going with UE4. I had a rough playable demo working via blueprint relatively quickly which impressed my team. But as I got more serious about what I wanted to do, I ran into problems. In UE3/UDK I was pretty handy just using Kismet and modifying Unrealscript. Blueprint, which is supposedly more powerful and versatile - should have met my needs - but to be honest, I find it cumbersome and convoluted in comparison to Kismet. Simple things like sharing a variable, typical math operations, or running a score up requires a lot more work and setup, and everything feels a lot more restrictive when it comes to flow control. I feel like I need to work backwards and know ahead of time exactly how to set something up. Following tutorial videos solved some problems, but when the videos themselves seem to not know why something had to be done a certain way (for instance having to create several nodes just to get a “Cast To” node to appear instead of just typing it in), it feels like a huge step backward from Kismet where never once did I have to follow a video to figure something out. Small problems would begin to pile up, and I would spend hours looking up solutions, which decimated my productivity. I would go to answerhub for solutions, but this only worked in the beginning. In the first month or so, my questions would get prompt attention and I’d have my problems resolved relatively quickly. I was not only pleased, I was impressed, and felt my $19/mo was more than well spent. The second month and beyond, it was a totally different story; I posted 3 times with no solution - and I’m not feeling much like spamming a board if there’s really no one listening. There could be dozens of questions I could be asking but if one simple one is given no attention after several attempts, didn’t give me any faith that something more complicated or critical would be answered down the line. I felt that being able to have some of my questions answered was a huge part of what I was paying the $19/month for, and it would be what I’d tell my fellow devs when encouraging a decision to switch “well, if you pay them by the month, they have to help you, right?”. I suppose I was mistaken that a subscription payment would be any guarantee of support attention.
There were more serious things such as engine crashes and blueprint data corruption. At one point the engine caused a bluescreen of death when I closed it, and some of the blueprint files I was working on turned into empty files full of zeroes, which was a pretty terrifying setback. I’m sure it sounds crazy… but the dev I talked to about it on answerhub could provide no insight into what had happened or how to avoid it other than that it shouldn’t happen.
In the end, what it boils down to is “am I definitely going to ship a game with this?”. The monthly fee we pay is a waste if we don’t actually ship something on it. Early on I had every confidence, but as the clunkiness of Blueprint became more apparent and odd things not working for no logical reason began to pile up - I became less productive and it has become a painful game of wondering if anyone will answer my answerhub questions or if my problems will be magically resolved in the next incremental update, and as the hope of either has faded, the $19 a month becomes a bleeding out and a detriment to my productivity. That’s money I could have spent on some plug-ins or better lunches. Its also time and effort that could have been spent developing on another engine that would ship the game, that won’t be carried over. I have every desire for Epic to get their cut of my game if it ships on the UE4, but if I am getting frustrated, having to pay for the engine being in a not-quite stable, not quite complete state where I can’t be confidently productive becomes a bigger burden as I wonder if “another month” will fix things. While I don’t feel like the monthly I’ve been paying for UE4 has been a total waste, I feel that if I continue at this point, it will be.
That being said, I am still very enthusiastic and supportive of the 5% pricing model that is being used for UE4, and sort of wish that I could apply that sort of deal to developing a game with UE3/UDK (I miss Kismet), but I know there isn’t much hope in that since all the attention will be on UE4 from here on out it seems. I sincerely hope that the engine will mature, blueprint will evolve to be more intuitive, and that answerhub will get staffed up to handle the volume - so I’ll be able to seriously consider using the engine again for a future project - its just that right now I need to be able to ship something quickly and reliably, and at this point there are too many stumbling blocks and unknowns. If you are a Epic developer or are another UE4 user, thank you for taking the time in reading this; hopefully this can be taken as feedback or inspire others to voice their opinions as well.