Yeah, I think my issue stems from one iteration of a seed I used to farm iron in the swamps I started several months ago. I started A recent version of that same seed (noting the case and letters) and it's very different. I think I went too long between iterations. makes sense, thanks guys!

You can use the site valheim-map.world to type in your key and it will show you the location, size and shape of every biome and item, so there shouldn't be any difference in the layout. Whether it goes down to individual stones, trees and mushrooms I don't know, but I assume you are talking about differences in the broader sense.

Originally posted by umop-apisdn:

Originally posted by Faceplant8:

I would love to see proof of any differences. I've done tests in the past and have seen absolutely no difference in anything.

Phase 1:
Make a new world, with whatever seed you would like.
Explore it for an hour or two, noting the locations, varieties, and counts of trees, mobs, etc.
Enter a burial chamber, note the layout.
Take note of the locations of metal deposits.
Take note of berry bush locations and numbers.
Take screenshots.

Wait a real-world week.

Phase 2:
Make a new world, using the same seed as the world created earlier.
Wander around, looking for the landmarks you noted during exploration of the previous world. Notice that the world created earlier is not entirely identical to the new world.

Tada! Proof that world-gen is only semi-deterministic.

I'm sure you saw my comment that I have done this before, so I'm not sure why you felt the need to state the obvious.

This is a link to my previous test: https://steamcommunity.com/app/892970/discussions/0/3496509339509393442/?tscn=1667873867#c3496509872205572846

I did note back then that the burial chamber was different, but didn't find any other differences.

As it turns out, I created a new world a few days ago, so I recreated it again today, and created another pair of screenshots from the same location in both worlds. I can't see a difference.

Here is a screenshot of the world created a few days ago: https://steamcommunity.com/sharedfiles/filedetails/?id=2943387920

This is around the same vew from the world created today:
https://steamcommunity.com/sharedfiles/filedetails/?id=2943388141

Please point out any differences that you see, and post any results that you have from similar tests.

Originally posted by Faceplant8:

Snip

Have you tried entering the same crypt in each world? You can actually test with a single world. Enter a crypt, explore a bit, leave, rollback to a previous backup and get into the same crypt. The layout will be different. Like you said "I tried it".

Yes a lot of the world will be identical with the same seed, but there are parts that use a random parameter on the fly when generating areas as they get discovered.

Originally posted by vinyblaster:

Originally posted by Faceplant8:

Snip

Have you tried entering the same crypt in each world? You can actually test with a single world. Enter a crypt, explore a bit, leave, rollback to a previous backup and get into the same crypt. The layout will be different. Like you said "I tried it".

Yes a lot of the world will be identical with the same seed, but there are parts that use a random parameter on the fly when generating areas as they get discovered.

Yes. That's what I stated in my post above that the crypts are different, but the vast majority is the same.

Originally posted by Faceplant8:

I would love to see proof of any differences. I've done tests in the past and have seen absolutely no difference in anything.

Phase 1:
Make a new world, with whatever seed you would like.
Explore it for an hour or two, noting the locations, varieties, and counts of trees, mobs, etc.
Enter a burial chamber, note the layout.
Take note of the locations of metal deposits.
Take note of berry bush locations and numbers.
Take screenshots.

Wait a real-world week.

Phase 2:
Make a new world, using the same seed as the world created earlier.
Wander around, looking for the landmarks you noted during exploration of the previous world. Notice that the world created earlier is not entirely identical to the new world.

Tada! Proof that world-gen is only semi-deterministic.

Some things are completely defined by the seed, for example location of biomes, location of crypts, etc.

However, the areas are only fully fleshed out when your character gets near them and that's where you get some randomness. For example, try saving the game, entering a crypt, reloading and entering again. The layout will most likely be different because the interior of the crypt is generated when you get in.

This is why the devs were telling people that as long as they had not entered a Mistland area before the update their map should be good to continue. This is also why online seed generators cannot show everything, like location of silver veins for example.

Originally posted by ManAnt:

The crypts and such are not always in the same place, Regardless of the seed you reuse. Rock formations and building placements also often are off from where they was when previously created..

Originally posted by umop-apisdn:

In my experience, using the same seed on two different days will result in slightly different results; it would appear the world generation is only semi-deterministic.

Its based on patches (afaik). Big patches change terrain generation. Small ones that include fixes to spawn locations etc will re-roll the objects on map, others will not.

So there can be long periods with maps not changing at all, depending on frequency and specific details of given patch. We just dont have good enough patch logs to be sure at a glance.

Part of the problem is that the maps are designed to be compatibile across multiple patches, so the generation can change mid play, and its unnoticable till you make another world.

This thread answers my six-year-old Rimworld questions.

Originally posted by umop-apisdn:

In my experience, using the same seed on two different days will result in slightly different results; it would appear the world generation is only semi-deterministic.

The overall look and feel of the terrain will be mostly the same, but mob spawning locations will differ slightly, as will the location and contents of instanced dungeons... never mind the trees, rocks, and other features.

The precise location and makeup of metal deposits appears to me to fall under "other features".

I would love to see proof of any differences. I've done tests in the past and have seen absolutely no difference in anything.

Version differences will make a difference, of course, and I did not try the tests on multiple days, but I don't see why the day would make a difference. If I create a world with a given seed today, and it's different if I create it again tomorrow, then the date is obviously being added to the seed, which just makes a new seed.

Any game detail that is being generated using something other than a single seed-based pseudo-random number generator would have to be both transmitted in real time between players in multiplayer and stored in the world database. Both of which make absolutely no sense for static content.

Mob spawns, changes (terrain modifications, buildings, etc), and visited locations, of course, have to be transmitted and saved.

Seeds always produce identical worlds. However over the course of early access development there were a few changes to the way worlds are being generated. That means a seed from two years ago will look different today.

In my experience, using the same seed on two different days will result in slightly different results; it would appear the world generation is only semi-deterministic.

The overall look and feel of the terrain will be mostly the same, but mob spawning locations will differ slightly, as will the location and contents of instanced dungeons... never mind the trees, rocks, and other features.

The precise location and makeup of metal deposits appears to me to fall under "other features".

The crypts and such are not always in the same place, Regardless of the seed you reuse. Rock formations and building placements also often are off from where they was when previously created..

Also pay attention to capitalization. The name should be exactly the same.

They differ between terrain altering patches, so mostly big updates. Otherwise they should be the same.