Slashdot Mirror
Building the Infinite Digital Universe of No Man's Sky
An anonymous reader writes: Hello Games is a small development studio, only employing 10 people. But they're building a game, No Man's Sky, that's enormous — effectively infinite. Its universe is procedurally generated, from the star systems down to individual species of plant and animal life. The engine running the game is impressively optimized. A planet's characteristics are not computed ahead of time — terrain and lifeforms are randomly generated on the fly as a player explores it. But, of course, that created a problem for the developers — how do they know their procedural generation algorithms don't create ridiculous life forms or geological formations? They solved that by writing AI bot software that explores the universe and captures brief videos, which are then converted to GIF format and posted on a feed the developers can review. The article goes into a bit more detail on how the procedural generation works, and how such a small studio can build such a big game.
Set it up so they just all upload to imgur. If a gif gets more downvoted, nuke it from the universe.
Definitely feel a Peter Molyneaux coming on - before you know it the hype will go so mad, you won't even notice that the game's actually been released, and then we'll find out it's as dull as hell as a game.
But aside from that, a team of 10 isn't exactly tiny. A lot better games have been written with a lot less people.
And front-page of Slashdot before release? I'm guessing at least one of those people works in marketing...
I look forward to playing the hell out of this game. And also seeing whether it's a failed promise, a glorious achievement, or both or somewhere inbetween.
Jesus Christ, it's terrible when somebody advocates "downvoting" as something of value.
Just look at reddit or Hacker News today, or even Slashdot, for example. The discussion there is horribly sterile, all thanks to their flawed moderation systems that encourage rampant censorship of any ideas that are unique or worthwhile. Comments expressing such ideas are near-instantly downvoted out of view.
That kind of a system fails even worse when the audience consists of people who are commonly labeled as "hipsters". These people have a very skewed perception of reality. They pretend to embrace "tolerance" and "acceptance", but they're often among the most strident crusaders against any sort of free thought and free expression.
We see the same with academic peer review, where the theory behind it is sound enough, but in practice it's just the entrenched players demonizing and shitting upon any challengers. Hell, we see the same with democratic political systems.
I don't see why your system would be any different. Unique game content would be quenched by the stupidity of those who seek to enforce total and unrelenting conformance with what they like.
...and I'm generally not interested in games. But this could turn me into a convert - the concept seems really awesome, and the sample video looked very cool.
Just now I've slipped off my armour of techno-jadedness, and I'm amazed at the wonders we humans are capable of creating when we're not busy engaging in pillaging, war, and petty bickering. Off-topic perhaps, but what the hell.
'The Economy' is a giant Ponzi scheme whose most pitiable suckers are the youngest among us and the yet-unborn.
I remember a text dungeon game in the 1980s that had a dynamically generated map. If you set the seed to the same value, you always got the same map.
Seemed kind of revolutionary in 1984 for some reason because you could have a huge map without actually having to create a huge map.
While I'm not really looking forward to this game specifically, I am glad to witness procedural content generation technologies advance as they have. From TES: Oblivion's foliage to Starforge's infinite terrain demo, things are becoming very interesting.
Back in my day, we had 800 planets that were procedurally generated, AND WE LIKED IT!
But really, the concept that you can procedurally generate an infinite universe is not that ground-breaking. I understand a lot of people are falling for the hype-train. And hey, it might be a really fun game. As long as there's something meaningful to do with all this PGC, and the game play holds water. But just because it's procedurally generated doesn't sell me on a game.
(Also, side-note, where the hell are the procedurally generated maps for the FPS genre? Why hasn't this happened yet?)
Thought this game might be using Outerra but no, it's not:
http://www.outerra.com/
http://www.moddb.com/engines/outerra-engine
Frontier Elite 2, for instance. Ken Musgrave literally wrote the book on procedural generation and is the brains behind MojoWorld, a procedural world generator that's great fun. If you liked Bryce back in the day, MojoWorld is Bryce on steroids.
Not knocking these guys at all, btw, it looks great. Just giving some background.
If you were me, you'd be good lookin'. - six string samurai
Reminds me of Elite of old past. While fun for a while, the similarities got boring and tedious pretty fast.
This is true for all current procedural games. Eventually there'll be enough similarities to make it boring.
Even a game as beautiful as this. Random critters and landscapes just ISN'T CONTENT.
It's a never-ending chase: Maybe if the gameplay become procedural as well, it might be interesting.
However, you'll have difficulty creating something with the longevity of L4D2, without narrowing down content and gameplay into something which is just right.
Balance and fun is hard to make procedural. Which is why I think this is a fool's chase in the most general case, but to be fair: intellectually stimulating to pursue.
Captcha: advert
42
needs more linux
US$0.02++
So they're making Spore 2? Or maybe Minecraft: Corporate Sellout edition? We've heard this song and dance before.
Of course, since I've been playing Roguelikes for decades, literally nothing in the hype train up there really gets my attention. Wake me when they have something with 1/10th the complexity of Dwarf Fortress.
I was fascinated by the idea of Spore when it was in development, and not so much when it was released.
Replays might be more interesting, but you're going to have to make a game pretty damned good for me to want to reply it -- and you're going to have to hope that the procedural generation the second time around makes for an interesting game.
If I play it once, I can't tell the difference between on-the-fly generation and static worlds.
It is stated "terrain and lifeforms are randomly generated on the fly as a player explores it" but this is completely false, and the developers have even said so in interviews. Basically, terrain and lifeforms are "evolved" from seed values. A given planet with the same seed values will always generate the same planet. It is not random.
From the summary:
"A planet's characteristics are not computed ahead of time — terrain and lifeforms are randomly generated on the fly as a player explores it."
From the article on Eurogamer (and reality):
"We do start everyone on a different place on the outside edge of the galaxy," says Murray, reiterating his grand statement from Sony's conference. "It is the same for everyone, but they are actually not just hundreds of miles apart, they could be billions of miles apart."
The game universe is the same for everyone. The devs are generating it and when they are happy with the way it turns out, that version will be the one they release for everyone to play in.
Procedural generation of outdoor scenes has been done for years. Decades, even. Works fine. Most of nature can be simulated with fractals, and basic terrain generators are simple. Speedtree turns out really nice trees and vegetation.
What's really tough is procedural generation of cities. There are programs that build a skyline, but so far, nobody has been able to procedurally generate a convincing city at high-detail level. There are systems that tried, like Introversion and Instant Architecture, but the sameness of the buildings makes for boring cities. It's easy to do this badly, but very hard to do it well.
It's not impossible. You'd need something like The Sims engine or SimCity, which grows cities over time in response to their occupants'' needs. That would be a big win. You could build something like GTA without an army of people constructing the real estate.
Those not agreeing w/ you below your post's just an example of assholes who abuse the easily cheated so-called "moderation" system here.
WANT.
Hopefully insurance salesman free this time. I told God last time...
There's a thought for you...
It's not just the idea that we may be a simulation, but we may be one where our creator isn't even aware of us.
I get how you can procedurally generate an environment on demand. What I don't get is how do you save the state of the procedurally generated environment after it's created, so that when you go back again, it still looks the same? Sure, everything is parameterized, so you're not saving nearly as much as you see, but if this is universe sized, and you get many thousands (or more) of people playing it for a while, wouldn't the size of the data set grow exceedingly large?
The cat won't even be a cat until somebody looks at it.
It does initially feel like Spore, but when Giantbomb were talking/hanging out with a some of the developers (E3 Day 1 GiantBombcast, at around 30min in or so), they pushed the devs about what the hell you do besides exploring and they didn't go too far and promise too much. They did mention some ideas that haven't been completely fleshed out yet: combat - space and planet-level, exploring (sharing? ugh), resource mining, ship upgrading. I would personally like it if they created some giant ship-design tree that would possibly be based on what your planet(s) had. But as they mentioned, these either weren't showable, or were just ideas they had, so it's probably best to wait it out and see if they can find some good gameplay mechanics and game goals. Otherwise it'll be "201X's Spore". (although I personally didn't mind Spore at the smaller kens; it's just that I remember my last hours of Spore were playing a frustrating planetary micromanagement sim. )
0- Eamonman Proud member of DNRC
that's enormous — effectively infinite
What does "effectively infinite" mean? Is it a) really, really, really big, so big that you'd never be able to explore more than a minuscule fraction of it in a lifetime (but not infinite) or is it b) infinite?
systemd is Roko's Basilisk.
Another space game that is being based on procedural generation is Infinity: Battlescape. The company behind it, I-Novae Studios, consists of a couple developers and a couple artists, many of whom are working part-time. The lead developer, Flavien Brebion, has been working on the thing part-time for the last decade. He's very good and very much the perfectionist, but hasn't gotten around to building an actual game yet (other than a fun little combat game intended to test networking code).
They have put out a few videos through the years. You'll note that they favor realism over the more cartoonish feel of No Man's Sky. In NMS, the fighter lifts off from a planet surface, travels what looks to be a couple miles and is out in space, looking at 'asteroids' and space battles. That's great for game mechanics, but it kills the sensation of realism for me. In contrast, the expectation is that in Infinity: Battlescape you've gotta fly 300,000km to get from the Earth's surface to the Moon's surface. Naturally, your ship will be quite peppy so that you don't need three days of real time to make the trip.
I-Novae Studios is shooting for a Kickstarter sometime soon.
Any time now.
I'm sure it's right around the corner.
I've always assumed this might be how the real universe works.
Atoms only exist upon detailed inspection. If noone is using a powerful microscope, no need to populate that detail.
The moon only physically existed once we landed on it and most of the stars will never need to physically exist.
It makes the whole "universe is a simulation" much more computationally feasible. It might even be able to explain
things like the double slit experiment and definitely helps explain why we seem to be alone.
Just a note that if you are a PC gamer, the game may not even release on that platform. It will initially release on PS4.
I'm bummed. For the longest time I thought this was a PC game.
Haven't seen Limit Theory mentioned yet. It's an infinite space sim, all of the universe is procedurally generated, and coded by a single guy. Looking darn pretty too!
Though, it's not down on a planet level, it's only in space.
/ The Arrow
"How lovely you are. So lovely in my straightjacket..." - Nny
Even if the game is bad... this kind of tech is much needed and could fuel much deeper gameplay down the road.
Starbound is a 2d side-scrolly infinite universe procedural game.
A couple weeks ago, I noticed a post by someone who had seen a randomly-generated mob of a specific type on a certain planet, taken it for a pet, and then subsequently lost all of his data, perhaps due to a patch, I forget. He wanted to find that exact pet again because he liked its looks, but in a very large (not technically infinite, but impossible to explore for one person before the heat death of the universe) game universe, the chances of him finding it again were very slim.
My suggestion was to write an AutoHotKey script and game mods to make the player invulnerable and run a ship on autopilot and explore the universe, then put that script into the cloud and run thousands of simultaneous copies. Then, post screenshots when mobs were found as Amazon Turk jobs to compare to the original screenshot of his lost pet.
5 bux says this game will flop badly the reason being that it will be primarily centered around exploration and most likely space combat rather than anything else. Exploration is nice and all as well as space combat, but both get boring real fast. I think a lot of people are pinning their hopes that it will be the next X game of sorts or perhaps something better than Star Citizen altogether, personally it's foolish thinking considering the videos and the plans the game maker has for it. They are just not taking it in the direction of economy, build stuff, alliances, etc.