Minecraft as a dedicated design tool

inlimbo

Newcomer
This is something close to me. I dream of writing it my sleep. As a programmer who cares less for the programming and more for the tools it can yield, it's a project I wish I'd made any sort of real progress toward in the last ten years. And, yeah, it's been spinning its web in my waking and unconscious mind for that long, with equal parts unbridled enthusiasm and Derek Smart-levels of self-delusion, well before Minecraft existed. But Minecraft is great inspiration. You can see that across genres and scenes right now. Problem is too many devs are trying to replicate its success wholesale, with minor tweaks for playability, while ignoring its greater potential as an entry-level 3D design tool.

3D game development as of now is necessarily decentralized. It takes groups of people in specialized roles to put together something like Assassin's Creed 3. 3D tools then necessarily cater to that workflow. It's a very piecemeal process, but it works if you have those connections and feel comfortable in that environment. It's not the only way to work, or even the best, however. Game design should be a very hands-on process in my mind. Designers should contribute to art, animation, world building and scripting in a direct fashion. Or at least they should have that option. In the days of 2D tile engines, this was very possible. In the days of early 3D, it was even still. But with 3D it's never been as basic and functional as it is with a good, feature-rich 2D tile engine. Minecraft turned that idea on its head, but it's still not all the way there.

Now consider Blockscape: here's a guy taking the basic build aspects of Minecraft and expanding them, giving them nuance, without sacrificing the fun and the ease of use which Minecraft engenders. But he's set his sights on the same ends Notch achieved - a contstruction set meta game, a sandbox mmo - and seemingly not much else. What he could have is a tool that could change the development landscape for indies looking toward 3D. Maybe the secretive "Phase 3" is something like that. Maybe not. But it's what I want. Accessible, but relatively freeform tile-based 3D engines are the future of independent development as I'd like to see it. If you take Minecraft as a bridge, expand its tile sets, give users simplified scripting tools and give them basic 3D model templates that can be easily modified, textured and animated in-engine (this is key), you'll see single persons putting together hugely ambitious 3D games.

Imagine putting together Super Mario 64 by yourself in six months? Or Ocarina? It's not far out of reach with the right tool. Might not look like something a team could put together in UDK, but it wouldn't matter if it played well. And, with current generation hardware, who says it couldn't look like that? Powerful tools with basic and direct interface could mean all the difference to the laypersons out there.

I'd really like to make this happen. What do you think? Have I burned through too much of my goodwill on this forum?
 
Last edited by a moderator:
I mean, there's Gameglobe out there and it's exciting, but it's still largely based around content authored outside the editor itself. I consider that a barrier to accessibility and to longevity. And it's not ideal if you're talking one or two people who have a vision that doesn't fit within the confines of some given prop set.
 
Think of it this way: you're working in a 3D level editor that's basically a lego construction set, much like Minecraft. Unlike Minecraft, you have options. You can define your blocks. You can define your switches/scripts. You have basic modeling, texturing and animation tools in-engine, in-game.

The idea is to change how a single person could build a game. In UDK you have kismet, you can build prefabs and align them to a grid and you can do some limited animation, but it's all very abstracted and it caters to a traditional game development pipeline. I want to change that environment. I want the editor and the game united, as it is in Minecraft, to cater to those people who prefer more direct, hands on work.

I want to do that because it's how I'd like to work.
 
Did you try the Cube engine?, it features real time level editing even during multiplayer and has been around for a decade or so.
I've never really played Minecraft but Cube was here first, the engine is kind of a hack with severe limitations but it ran very fast and its nature made it lack other limitations (no map compilation, and one of the features was "real time static lighting" so it didn't look like total crap even without precomputed lightmaps)

Cube 2 is basically the same thing but with modern stuff (pixel shaders etc.) and it says it's based on an octree, whereas the first one really is little more than a heightmap on top of a grid of squares.
 
Man, I forgot all about Sauerbraten. Cube 2 is insanely impressive, probably moreso than Blockscape, but it still lacks the scripting and animation stuff I'd like to see, since it's geared toward arena play. It's as good a starting/reference point as any, though.
 
So your post is "I'm going to make a game you can edit in real time" your not asking for anything in particular.
 
Think of it this way: you're working in a 3D level editor that's basically a lego construction set, much like Minecraft. Unlike Minecraft, you have options. You can define your blocks. You can define your switches/scripts. You have basic modeling, texturing and animation tools in-engine, in-game.

The idea is to change how a single person could build a game. In UDK you have kismet, you can build prefabs and align them to a grid and you can do some limited animation, but it's all very abstracted and it caters to a traditional game development pipeline. I want to change that environment. I want the editor and the game united, as it is in Minecraft, to cater to those people who prefer more direct, hands on work.

I want to do that because it's how I'd like to work.


I guess producing such tools isn't easy, now you need decent and easy to use modeling software right inside your game and all that seamless scripting stuff. Everything needs to be the most simple, the most powerful while never crashing and being low on resources and showing real time results. I can only guess this would be hard to write.
 
It won't be easy, but it is possible. If i'm being too vague, it's because I wanted Minecraft or Blockscape or Cube 2 to stand is as reference. And I'm not asking for the world here. An editor like this should be allowed to be necessarily inefficient, performance-wise, if it needs to be. You want performance and you scale your world and your assets according to hardware.

I just long for a unified work environment. Look at Source Filmmaker or Unity 4's Mechanim toolset. If you could scale those tools back, give users a straightforward frontend where they could plug bones/IK rigs into whatever they've modeled in-engine and allow them to store and replay keyframes all at once, you could achieve similar results without having to implement all the luxuries you'd find in a complete animation package. I have my eyes on relatively low poly modeling here, anyway. Give users basic extrude, subdivision and bevel tools and you're cooking with gas. And since they'd also have access to tiles and blocks (which could then be modified as necessary), it would allow them to iterate quickly.

When I say "simple but powerful" I don't mean the most simple and the most powerful. That's a losing game. Consider it necessarily vague wording instead of abject hubris on my part. I'm willing to be realistic.
 
Last edited by a moderator:
I try to be realistic too, I'm very vague as well with no dev experience to speak of. Just an intuition that designing software that is easy, forgiving and does everything is probably hard. I always remember when the Winamp devs were creating that brand new audio player that would be able to do anything and would be infinitely and easily customisable and skinnable.. And they ended up with Winamp 3, an incredible failure, they had to scrap most of it.

Then, revolutions never happen because everyone knows it's impossible. Until someone comes who doesn't know it's impossible, and does it anyway :p
 
I try to be realistic too, I'm very vague as well with no dev experience to speak of. Just an intuition that designing software that is easy, forgiving and does everything is probably hard. I always remember when the Winamp devs were creating that brand new audio player that would be able to do anything and would be infinitely and easily customisable and skinnable.. And they ended up with Winamp 3, an incredible failure, they had to scrap most of it.

Then, revolutions never happen because everyone knows it's impossible. Until someone comes who doesn't know it's impossible, and does it anyway :p
Wow nice Marcel Pagnol quote, unexpected here :)
 
Well, you know what MLK always said: If I'm gonna get high, might as well get as high as possible.
 
Last edited by a moderator:
I couldn't find a Marcel Pagnol quote, but stumbled upon a reference to this one ("Should he become Shaw’s reasonable man?")

The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man.
George Bernard Shaw, Man and Superman (1903) "Maxims for Revolutionists"
 
I'm not interested in great men. I don't dream this as a sweeping sea change this industry requires. That lies in economic and managerial restructuring and reassessments of long term company goals. I'd rather think of this as a potent alternative to classical game development. The stubbornness is ancillary, not a driving force. If I could realize this in any way that made 3D game development a designer-driven and relatively unified process that could be undertaken on an individual basis, then I'd consider that a win, limitations and all.
 

I missed this footage before it got pulled from GameTrailers back in July (as did most people), but this - this is the dream. Of course the chances of putting together something as sophisticated as this on my own are slim to none, unless UE4 proves otherwise, but it's nice to dream. And I guess I could imagine doing a leaner equivalent.

edit: there's little scripting in there and you're not rigging characters or setting up game logic in fortnite - at least not yet - but, as far as building from presets/tiles with seemingly few structural limitations goes, it's impressive stuff.
 
Last edited by a moderator:
Back
Top