Install Steam
login
|
language
简体中文 (Simplified Chinese)
繁體中文 (Traditional Chinese)
日本語 (Japanese)
한국어 (Korean)
ไทย (Thai)
Български (Bulgarian)
Čeština (Czech)
Dansk (Danish)
Deutsch (German)
Español - España (Spanish - Spain)
Español - Latinoamérica (Spanish - Latin America)
Ελληνικά (Greek)
Français (French)
Italiano (Italian)
Bahasa Indonesia (Indonesian)
Magyar (Hungarian)
Nederlands (Dutch)
Norsk (Norwegian)
Polski (Polish)
Português (Portuguese - Portugal)
Português - Brasil (Portuguese - Brazil)
Română (Romanian)
Русский (Russian)
Suomi (Finnish)
Svenska (Swedish)
Türkçe (Turkish)
Tiếng Việt (Vietnamese)
Українська (Ukrainian)
Report a translation problem
Oh wow, so are you saying that this is just like a Unity limitation?
To quote Dave Newson's site :
Read Here [davenewson.com]
http://davenewson.com/posts/2013/unity-coordinates-and-scales.html
As posted in the answer here [stackoverflow.com];
https://stackoverflow.com/questions/49752582/unity-coordinate-limitations-and-its-impact
They use Perlin Noise with a different math algorithm per biome. The seed is used for random offsets in those algorithms.
Perlin noise is known to having problem the larger the offset, why 10000 is a good mark to stop.
Want a walk-through, check out this series of tutorial videos (using Unity): Procedural Landmass Generation
Wait - what? I play in three worlds (my world a friend's world and another friend's world) and all three have plenty of continents; in all three total landmass far exceeds watermass.
This is pretty crazy to me, but if this is true then I start to understand why the total landmass is so small. If you can't provide a seed then all you can do is offset the sampling locations to get a different result. This is indeed what Dvoid seems to have done. Also in this scenario it is plausible that the quality starts to degrade if you sample too far from the origin.
They should have used a custom noise function that can be seeded instead if this is true. Its 10 lines of code or something. I guess its too late now, we literally have a hundred million savegames out there that all depend on the built-in noise function. Oh well. At least this design decision starts to make some sense to me now, assuming this speculation turns out to be correct.
Offset is random from seed to get a delta x and delta y value.
Then they loop for x and y coordinate values, add/multiply depending on algorithm used the offsets and other values (Mathf.PerlinNoise(generated_x * preset_floats, generated_y * preset_floats) * (more Perlin noise calculation)). It is more involved than Sebastian Lague's example for a base height map, but still similar concept. Then biome heightmaps are more Perlin noise with different add/multipliers for offsets and coordinates as well.
In the end, a lot of games use Perlin noise. There are other ways of course (NMS is a mixture, including Perlin noise. Is a YT video explaining it), but if they start with Perlin noise, then they likely won't change. But as they have mentioned Ashlands will be differently calculated than it used to, hence why you should start a new map. Or at least hope the zones around Ashland wasn't calculated if you got near enough (as soon a new zone if generated, it stays like that in the world save "database" (binary) file.
But it is Early Access, nothing stops them from changing. In the end, I think Valheim does an excellent work on generation. And I feel OP is just yelling foul for no good reasons.
When Mistlands was introduced they changed the terrain generation a little bit. They tried to patch people's saves but if you were unlucky and you had built near the old Mistlands your buildings might have collapsed because the ground was not where it used to be etc. If they made a total change of terrain gen everything would burn in flames. Feels like they won't be able to do a drastic change at this point.
Valheim's world gen is pretty decent, no disagreement there. Not sure if I would have ever noticed the fact that the landmass is recycled between most seeds, had someone not mentioned it in the Discord. Its not a huge issue for sure, but it is a little disappointing that the world is not unique when you start a new game. Millions of people have seen the same islands and continents (albeit with different biome distribution). Depending on how much you have explored maybe you have too. Not game breaking or anything, just a little unfortunate.
Well, on the Ashland world recreation, it really is Early Access and people should read what it means. Things can change. Other games have wipes a lot of times, due to those reasons. I think Iron Gate has been nice for a long time already. It had to be expected that a "wipe" would have to happen. Especially those that have explored the whole world already. There it is harder to know what to "wipe' and what not. Either way, some people will get mad no matter what
And the recycled, well. Yeah, but you can also see that in other games. Hell, Starfield is hugely disappointing in how much they recycle.
In the end, I think people shouldn't focus on that. Is the game fun? Yes. Is it re-playable? In those cases, maybe take breaks, so it doesn't feel so obvious :)
Making games is hard.
The original German quote is "People who love laws and sausages shouldn't watch either one being made."
TL/DR: However it's done, it works for me.
This is where I feel DLC is appropriate. The number of different random buildings or unique rocks doesn't affect gameplay. Not worth it to Coffee Stain or Iron Gate to pay devs to create those. But there have been over 12 million copies of Valheim sold. If a million players want more models, then at $4.99 or so, that's ~$4 million in possible revenue.
I also play a game called "Railroads Online!". When you're laying track bed, the skin/pattern repeats. There just happens to be this one distinctive kidney-shaped rock right in the center of the skin. Players love this rock. To lay track on the roadbed smoothly, just plop the spline points on said kidney-shaped rock, and the tracks are both tidy and centered on the roadbed. Praise the kidney-shaped rock!
That's the entire point of constructive criticism. So we can tell them what we want improved and then they can do it.