Hey everybody, I am curious, what your biggest pains are in the game design and development process. What things are repetitive or take a long time? What problem do you wish there was a solution for, but there isn’t? Is there any area of Unity or game dev where there is little training and tutorials to be found? What tools (editor extensions) would you find most helpful? Do you find yourself wishing there was an ebook written on something like “How to create a _____ game in Unity”?
My main gripe is waiting for my pc to do something before I can continue, i end up talking on the internet too much
I feel that most lower level topics such as Shaders, Edit Extensions, “How to make a top down shooter” are already very well covered.
What I personally am interested in seeing is a much higher level understanding of how best to structure a Unity code project. I’m talking about design patterns, architectural structure of your project, when to use interfaces, applying good object-oriented design principles, and in general how things connect together to bring a really nice, clean, easy to understand, expandable and optimised code.
Lightmapping is something I absolutely dread 
Especially when project nears completion and you need to bake higher quality maps, makes iteration time on last minute changes and tweaks skyrocket.
Good looking artwork, takes way to long. Always need more and making it look good is no easy task!.
You can re-use code, it’s hard to re-use assets and get away with it (On a continual basis).
Also rapidly having to relearn something when new shiny tech comes out, I’ve spent so long with diffuse the thought of PBR makes me cry slightly… Even though once I’d figured it out it wasn’t that bad. But still, can’t we all stick at ID Tech 1 and just be happy?
@ShadowK
What’s PBR?
It’s Physically based rendering / shading mate. Good example here:
http://blogs.unity3d.com/2014/03/20/physically-based-shading-upcoming-in-unity-5-0/
Ah thought so (first time i hear some one refer to it as rendering instead of shading), anyways google said it’s professional bull riding, so… That’s why i asked XD
I’m a recent initiate to PBR and I absolutely love it, best thing to come out of realtime comp graphics in a while. Only problem is the getting the i for the ibl. Roll on next version of skyshop
bump
For me along the course of my 3 year part time indie project its been finding a damn good UI artist that I can be satisfied with!
Being a programmer, not having any reliable art direction has been not only a huge waste of my time, but extremely expensive too.
I am now on my 6th UI artist for my game, I’ve given him the go ahead to re-do everything from scratch.
I found a good artist for my tracks on the 3rd attempt, so happy with that.
I can’t imagine how frustrating it must be though for people who battle to find good programmers.
Finishing.
and animation. I HATE ANIMATION. Mechanim did not improve this situation.
Out of curiosity, what issues are you having with Mechanim?
For me it would have to be the glaring inconsistencies in cross platform asset creation, mostly when it comes to shaders. Unless you know how to program advanced shaders from the ground up, you will be in for a world of trouble making things scale well while also having a style to it. I own Shader Forge and i love that extension to death, it covers a huge area Unity have never been very good at, but these Shaders will not scale well to OpenGL from my experience.
Easily scripting events is also a real pain if you want to have moving elements and such. I have bought uSequencer for this and while it works pretty well, it is just such as far cry from what Unreal Engine offers.
Lack of instant transitions really bugged me… can do it with curves, but the documentation on curves is s*** at best, and I dont really feel like having additional animation parameters just to achieve something that should be available without.
I wanted to make it so my animations were automatic based on the character state, but instead, to get the result I wanted, I’ve had to put in place a system that specifically tells the controller which animation sequence to play, which I have to fine tune to make sure it looks good, which is all controlled by the total amount of time I want an animation sequence to play.
Its a clumsy way to do it, which I despise.
Im sure after spending more time than I like on it, I could probably get it to work better, but I really hate animations, so was hoping mechanim would make it simpler to get smooth transitions…
I could totally use that too. I’m constantly showing off some of my code, and having people tell me “add this(duh)”. “do this”, etc. There really need to be a free book of unity standards and good coding etiquette.
I found Mechanim to be a little helpful, in the sense of being able to use animations on multiple characters. But bad in the sense that it can take a lot of work to setup the rigging and the animation states. I would like to see more documentation and plugins to help you do that faster.
IMO, the only decent feature of mechanim is re-targetting.
Mecanim. For the same reasons that JamesLeeNZ gave + infinitely more reasons not given and or forgotten. The only good thing that anyone ever says about Mecanim is the ability to retarget animations (limited by design to only humaniod rigs). Something of which I am sure Legacy can do better.
Ya I’d have to agree with the whole mecanim thing. I think it’s overall design is good, but it’s got so many damn niggles I end up swearing at it half the time.
That and root motion. I was resistant to the idea, but it ends up making things feel so solid and consistent I really love the results. In my current game prototype the root-motion driven characters are far and away better than the others.
Having said that, for small teams that only presents an advantage if you’ve got easy access to a good animator and/or access to a good motion library, because some tweaking has to happen in the animation. But if that option is available to you the results are great.
I am liking the rest of it too, since it’s something that I partially made myself, and having a better system out-of-the-box is a huge win, but… some of the downsides are (or were?) pretty painful. There seems to be some use cases it misses, and there’s a lot of low level/data access that we don’t get, so we can’t cover those ourselves.
While the legacy system is cumbersome to use it does at least let us to tell the system to do exactly whatever we want. Pulling numbers from thin air, Mecanim makes 90% of things far easier but the rest far harder.
Having said that, it does seem to be improving, and I can absolutely see the benefit in releasing stuff before it’s finished so that a) real world developer feedback can guide it the rest of the way and b) people can benefit from the use cases where it’s already good enough.
Aside from access to data, my main gripe with it is stability. I’ve had a few cases now where I spent time looking for bugs in my own code only to find that something was a glitch in Mecanim and deleting then re-creating the related state fixed it. I’ve also seen twice a glitch where transitions from the Any State get corrupted/lost. It’s still an overall time saver compared to legacy, by a long shot, but the idea of losing work makes me feel uneasy sometimes. Thankfully there’s always version control.