Question/Poll to the steemit community... with regards to the Unity Tutorials... please choose
I have a few directions I am interested in going with these tutorials. I would like to put the question to all of you out there in which way you want me to go. I am going to give you a list of larger multi-tutorial project ideas that I could cover. Please comment on which one you want me to pursue. You can also optionally up vote a suggestion someone else already made.
I would prefer to work on things that you the community are most interested in from these choices:
Choice 1: 2D Game Using Unity Sprite System
Design: Easy
Coding: Easy
Concept: Using CC0 public domain art show the basics of using the built in sprite system. There are quite a few tutorials out there that already do this.
Choice 2: 2D Game using Custom Sprite System
Design: Medium
Coding: Medium
Concept: Using CC0 public art make a system that builds texture atlases on the fly from art we supply which makes it easily extensible. Build our own sprite system to support these atlases. Benefit the sprite system could support multiple texture types so unlike standard sprites you could use normal, occlusion, specular, emission, height maps, etc to actually make sprites that interact with the Unity lighting system. This would teach quite a few cool concepts, but will span more tutorials and delve into code more than the other. There isn't a tutorial like this out on the internet that I am aware of.
Choice 3: Voxel Cube world style project
Design: Medium
Coding: Challenging
Concept: Taking the minecraft block world version of voxel style. Supporting cubes with each side being a different texture, building texture atlas on the fly similar to the 2D project above. Eventually delving into Perlin noise, world building, chunking, and if we go far enough potentially threading. This is a more challenging style tutorial than it may seem. There are tutorials out there that cover a lot of this. I will likely approach it different than they do, but this won't be a completely unique idea.
Choice 4: Turing Type AI program
Design: Easy
Coding: Challenging
Concept: Make a text based program that parses input from the player and attempts to mimic and have an intelligent conversation with the player. Google: Turing Test.
How far I go on those tutorials really would depend upon steemit community interest, and my interest level.
Let me know which way you would like to go.
Whichever you ask me to do it will take me a little bit to get the initial tutorial together. I have ideas for all three of them, but I don't want to do a build and concentrate on them until I know which one you are most interested in.
Results of the Vote
1 for Voxel, 2 for Turing.
I'll focus on Turing Test style tutorials first. I'll do voxel tutorials later.
Lately I've been hearing good things about Godot, an MIT licensed game engine similar to Unity. Would you consider learning Godot enough to write a tutorial or two?
I am sure I would consider learning Godot sometime in the future, but not at the moment. I am also leaving the door open for other people to write tutorials. I've tried to encourage some people that use Unreal 4 to do so. There may be others. Yet learning a new engine takes quite some time to become proficient at it, and I am teaching Unity tutorials because that is what I am using in my own projects. I do not have intentions of switching my projects to another engine at this time. Far too much time, and money invested in them. Though in the future I'd certainly give it a look.
I did take a peek at godot just now. I likely will mess with it some in the not so distant future.
Both choice 3 and 4 sound interesting to me. But if all the AI does is mimic the user I might have to go with 3. I do have interest on how these smart AI's work on though
It would try to act like a person. How smart we can get it to be really depends on us. A truly smart AI is a massive amount of code and would not fit into a few blog posts. Someone could write a large book on that subject. I could give the foundations though upon which people could build upon and experiment on their own. We would start working along the lines of Eliza, but hopefully better. Add in some simple logic type system, and then run with it and see where we can go with it.
That one is a pretty easy one for individual people to tweak and create their own systems once they see how it works. It certainly would not be true smart AI. At some point if we could make it reach out and do web searches and attempt to find answers that way that'd be pretty cool, but I don't know if we'd ever get to that point or not.
Voxel stuff is pretty straight forward with what that would be.
I'm torn between the choices but my vote goes for AI-stuff then.
How about a Mahjong clone with all the altcoin symbols?
First version could be 2D on one plane, then you would extend to 3D.
If I do the first choice I'll likely make some weird variant of something similar to Desktop Tower Defense. If I do the second one I'd likely work towards some roguelike. The third one getting the voxel world working the way we want it to work would be the goal, what game people wanted to make out of it after them would be totally up to them. Fourth... is an odd puppy. I suspect that one would start out as my initial seed and grow into some weird thing we made together.
I'm visioning Watson like AI being given to every born child to feed theirthirst for knowledge so communication via audio would be nice to see if possible
Audio communication would require using an asset to convert audio to text for it to parse. That is something anyone should be able to do if they have access to such an asset after they know what to do with the text. The scope of my project would be simply how to process the text and attempt to do useful things with it and make it come across as human as we can with the limits of some tutorials. A Watson level AI would be a large book, maybe several. The audio processing would also be quite large. Though making something as I described elsewhere is doable, and then adding on text to audio and audio to text plugins would be all you would need at that point.
I don't want to make an existing game. I get pretty bored doing that. The techniques I show you should be sufficient for you to do something like that though. I'm not too fond of remaking games people have already done. :)
I will give your idea some thought though. That'd likely be pretty easy to make. :) And easy for tutorials is good. I do tend to aim pretty high a lot of the time. A desktop tower defense type thing likely wouldn't be terribly hard either, but likely more involved than mahjong.
Well it's up to you, do what you like to do! I would also probably like to see how 3) is done.
So far I have 1 vote for choice 4
Collage I made for the start of the steemit blog post on this topic here.
and 1 vote for choice 3
Those are from screenshots of some of my projects of this type.
Hopefully we get some more input. I likely will do both of those tutorials at some point I am just trying to determine which to do now.
My vote is for number 4