I thought since Ivy Bridge Intel CPUs have had hardware RNG capability. Is this not good enough for your work? You could get a fucking crapload of random numbers in a short period of time with this.
https://en.wikipedia.org/wiki/RdRand
No Man's Sky [PC]
- Thread starter Malo
- Start date
https://en.wikipedia.org/wiki/RdRand
Oh noes. That would.. uhh.. actually absolutely no difference. Unless you're some kind of manwolf who likes to mark his territory by digitally pissing your SteamID over digital worlds that there is very little chance any person would every see.
That said I've run across a neighbouring system discovered by "Microsoft Infopath". All technology in the system was running Windows Vista and everything was on fire.
I liked the idea that the player can leave something. Now the naming feels senselessly when you can name elements which will be deleted later. Then they shouldn't even implemented it.
As far as I would say the probability is not too low to see an already awkward territory. Finally, two players even managed to visit the same planet on the first day at the same time.
Last edited:
If the need for randomness is related to encryption, then per that link it may not be sufficient.
For No Man's Sky, since the procedural generator's output also needs to be consistent for everyone despite on a planet zero cross-communication, it can be undesirable since two users would generate inconsistent results for the same planet.
As a seed for the algorithm in the 18 Quintillion worlds that are "unique", it can also be undesirable:
http://dilbert.com/strip/2001-10-25
I guess but all the negative stuff in the wiki article looked like the typical linux crowd bashing anything that isn't open and free. In reality I seriously doubt you'd have any problem using Intel's RNG even for encryption.
How does one even achieve a truly *random* number. I would think true randomness can only occur at the quantum level. And even then what if the Matrix machine that runs our universe isn't any better than Intel's implementation? Lazy devs and all.
Last edited:
D
Deleted member 11852
Guest
We define randomness by the volume of sample data before distinct patterns of repeatedness are detected. Things on a small scale can look random, but multiplied a billion times you see the patterns. This has always been a problem when you need a lot of random numbers. We use a variety of sources and scramble that data in non-random but non-predictable ways.
I hope Hello Games share some of the workings of NMS because no matter what you think of the game, the technology is very interesting.
I hope Hello Games share some of the workings of NMS because no matter what you think of the game, the technology is very interesting.
I would assume that if you wanted the reduction of repetition in selections (ie. the appearance of a more "random" distribution) then you would actually want something that's not completely random. It reminds me of when implementing sampling patterns for path tracing - you're going to get a much more visibly even convergence (less clumping of over/under-sampled regions) early on by biasing your sample selection in some way (stratified selections).
If there's a million people un-evenly distributed throughout the galaxy (presumably they would all be starting roughly equidistant from the center), and most are making their way towards the center, and most are discovering 10-100 planets during the course of their time playing, then you're not talking about a one in a quintillion chance of seeing any repetition within the entire group. On top of that, I would think that you would actually want to ensure that players are crossing paths to some small degree otherwise there's not much point to uploading your named discoveries to a central server to be shared out.
If there's a million people un-evenly distributed throughout the galaxy (presumably they would all be starting roughly equidistant from the center), and most are making their way towards the center, and most are discovering 10-100 planets during the course of their time playing, then you're not talking about a one in a quintillion chance of seeing any repetition within the entire group. On top of that, I would think that you would actually want to ensure that players are crossing paths to some small degree otherwise there's not much point to uploading your named discoveries to a central server to be shared out.
D
Deleted member 11852
Guest
I was thinking the reverse!
Take data as random as you can get and have the algorithm robustly eliminate ludicrous output like planets that only have aliens with 20 eyes. Without knowing how the algorithm works it's hard to say but I would like to know what the their algorithm is censoring.I think the tendency to have our intuition lead us astray on matters dealing with statistics and larger number sets probably muddies the waters a bit too much - the 'Birthday Problem' being one that seems relevant here. Unfortunately not only are we dealing with faulty intuitions but also an untrustworthy and/or unclear source of information, so trying to peal apart why certain phenomena are emerging is probably not going to produce anything of substance.
We don't know to what degree their game is actually a simulation (even in the most basic sense where, say, all the star systems have a defined relative arrangement such that paths remain constant across all players and the galaxy has a defined and constant geometry.) Considering how little of the game hinges on the larger structures of system/planet persistence and galactic simulation (minimal meaningful world persistence, no multiplayer, no waypoints, no need/focus on back tracking, a strong disincentive for multi-system travel (fuel cost), etc), I think it would be most likely that there's more sleight of hand than simulation in the game.
We don't know to what degree their game is actually a simulation (even in the most basic sense where, say, all the star systems have a defined relative arrangement such that paths remain constant across all players and the galaxy has a defined and constant geometry.) Considering how little of the game hinges on the larger structures of system/planet persistence and galactic simulation (minimal meaningful world persistence, no multiplayer, no waypoints, no need/focus on back tracking, a strong disincentive for multi-system travel (fuel cost), etc), I think it would be most likely that there's more sleight of hand than simulation in the game.
Yep, if the galaxy actually has some sort of consistent geometry with a center, then it would mean that the starting areas would be a shell that exists around that center. So instead of a volume of your total set as potential starting locations you'd instead have something closer to an area subset. Couple that with the large user numbers of a very front-loaded product launch and the weird 'Birthday Problem' math and it doesn't seem that crazy.
Can you talk more about this? Different stars need different warp drives?
Also, from where I started the game, there were a massive pile of stars "behind me." I am curious as to how an origin point is decided.
D
Deleted member 11852
Guest
I've just the one but assuming reactors are like all the other technology, there individual reactors ranges will stack for a maximum range. That's how weapon mods work, anyway.
Yeah, all upgrades work the same way. I made some tests when I rebuilt all my warp drives... the upgrades stack and it raises the maximum distance you can do per warp cell. You also get an additional boost by placing all you warp upgrades touching each other (you see this on the circular graph on the right, the shaded part is that bonus). Max I can do is 1650 light years with all three upgrades touching each other.
It took me quite a while to get the Theta blueprint, I don't know if it's random or part of the story.
The Blue stars (O and B class) tell you that you need Theta, but it didn't work for me, it needed both Tau and Theta upgrades. Not sure if they fixed it, there's been many patches since.
It took me quite a while to get the Theta blueprint, I don't know if it's random or part of the story.
The Blue stars (O and B class) tell you that you need Theta, but it didn't work for me, it needed both Tau and Theta upgrades. Not sure if they fixed it, there's been many patches since.
Silent_Buddha
Legend
And the people I've seen comparing the two, more often than not, say that Spore is a better game and that it isn't even that close. I found that pretty shocking considering most of those people thought Spore was a huge letdown and not a very good game.
For many it just came down to the creature variety and quality. It's just so much higher in Spore. I'm finding that many people are starting to see basically repeats of planets with minor differences to planets they've visited before. Creatures that are basically duplicated on a plethora of planets with little to no change between them.
But at least the games Molyneux released were arguably really good games even without everything he promised. And at least he didn't try to dodge around questions of missing features. If they weren't there, they weren't there.
Contrast that to Sean Murray who still wouldn't answer the question of whether there was the potential for multiplayer in the game days/weeks after the game launched. He'd answer with vague dodge answers along the lines of - The galaxy is so large that the chance of meeting another player is so slim that you should expect to never meet another player in game. When presented with a case of 2 players actually going to the same planet but being unable to see each other all he had to say was basically...that's amazing that in a galaxy so large that 2 players arrived at the same destination so quickly. And then proceeded to not answer the question about whether the game actually had any multiplayer code.
I probably wouldn't rate him lower than Molyneux if he'd just come out and say, something along the lines of "Nope, that feature unfortunately never made it into the game." when questioned about various features that didn't make it into the game. But no, that might turn people off his game, so he plays coy with answers. It's to the point where a lot of people just don't believe anything he says anymore.
Regards,
SB
Last edited:
The servers seem to be the issue with the name saving problem. Players could still give animals and plants names.
Interesting mod: http://nomansskymods.com/mods/bigthings-1-2/
Interesting mod: http://nomansskymods.com/mods/bigthings-1-2/
Last edited:
Similar threads
- Locked
