Truly infinite worlds (poll)

Would you like to have truly infinite worlds in Minetest?

Yes
60
56%
No
25
23%
That depends/difficult to say
23
21%
 
Total votes: 108

User avatar
voxelproof
Member
Posts: 1087
Joined: Sat Aug 05, 2017 08:13
Location: Europe

Re: Truly infinite worlds (poll)

by voxelproof » Post

twoelk wrote:LOL, a poll to declare infinity mappable, yeah sure
I definitely see here a true understanding of the concept of infinity :-D
Practically infinite worlds are something that already exists in the gaming. SpaceEngine mentioned above is only one example; in the voxel sandboxes genre Minecraft is also practically infinite. It's just a shortcut to express some psychological feature of the game, please do not take it literally.
To miss the joy is to miss all. Robert Louis Stevenson

User avatar
Dokimi
Member
Posts: 227
Joined: Mon Sep 19, 2016 03:52
GitHub: DokimiCU

Re: Truly infinite worlds (poll)

by Dokimi » Post

voxelproof wrote:
I think that the horizontal size of Minecraft would do.
Had to look that up... and yeah... multiple times larger than Earth... infinite... close enough!

(I know nothing about the technical stuff... why the big difference vs Minetest?)

voxelproof wrote:Yes, and this in turn depends on the type of gameplay... to say nothing of the capacity of such a map to host real terrain features.
Both good points.

Game style is very important. For building... current maps can fit a mega-city, no issues. Those people should be happy!

For exploration... Real world - walking you might cover 15-30km in a day (in-game walk speed seems about right). So current maps are a challenge to cross, but not an overwhelming challenge... A committed explorer could eventually explore the whole thing (it is the size of a small island).

Terrain size... Helpfully Minetest does make things feel 10x bigger. So a compelling kind of realism can be done currently. Still, for a realistic map you are limited to 1 focal landform, at the small end of the size range. Maybe 2-10x bigger would allow realistically large features?

Gameplay... I guess we wont know what a realistic Mt Everest map would be like until someone makes it! There is actually value simply in knowing that such mega-scale features are possible in a map. It adds to the ambience... the "wow" factor... even if only one nutter ever bothers to spend three wks climbing it! :-)
voxelproof wrote: Maybe you're right. More diversity in the default content is certainly needed,...
It's just little details really. A realistic plain can be 10s to 100s of km wide. But a new species of flower here, a unique rock formation there... that is what saves this kind of map from being a boring hellhole. For the current map size, Minetest currently has the bare minimum needed to pull that trick off.

User avatar
voxelproof
Member
Posts: 1087
Joined: Sat Aug 05, 2017 08:13
Location: Europe

Re: Truly infinite worlds (poll)

by voxelproof » Post

Dokimi wrote: (I know nothing about the technical stuff... why the big difference vs Minetest?)
It's all about the data storage. Probably.

My point is that writing a code which renders good terrain surface over huge area extending to the range of long integers is banal. A function like, say

y=int(500*(sin(.001*x)+sin(.001*z) ))

is computed in miliseconds and so the time needed to render corresponding blocks is that of a lightning bolt. This is how the things are done in Space Engine or other games based on polygon graphics. The real issue is that in voxel sandboxes you need to have also the earth's interior, all the layers of dirt, rocks, different ores, caves, etc. But this as I believe should be possible to be done in a similar way. I know that this would probably mean changing the present mechanics for land generation, however I maintain that this is possible to have fast and efficient engine for huge terrain area.

Whether this will be ever done in Minetest is as I see quite different matter, probably there are no plans to go beyond the present state of map generators. But, who knows, maybe someone in the future will make a fork which will please more demanding users :)

Game style is very important. For building... current maps can fit a mega-city, no issues.
Not sure, e.g. Adelaide in Australia stretches N-S for more than 50 miles and it's definitely not a mega-city :) Nevertheless there are no larger projects created in Minecraft than probably 5x5 km (Westeros) so in this sense you're perfectly right. There's no need for larger terrain just for building.
Terrain size... Helpfully Minetest does make things feel 10x bigger. So a compelling kind of realism can be done currently. Still, for a realistic map you are limited to 1 focal landform, at the small end of the size range. Maybe 2-10x bigger would allow realistically large features?
Frankly I like the present size, it makes some maps to look like huge gardens and even mountains in some maps, however not the size of real ones, are of compelling beauty . But wouldn't it be interesting to have some new mapgens and see really gorgeous worlds? Why to stop at the small scale when there awaits whole lot of things that can be done and provide yet more fun? Maybe such engagement incited by this kind of upgrade would be too costly for AAA producers? ;-)
A realistic plain can be 10s to 100s of km wide. But a new species of flower here, a unique rock formation there... that is what saves this kind of map from being a boring hellhole. For the current map size, Minetest currently has the bare minimum needed to pull that trick off.
Exactly! Glad to see somebody has so similar view on this aspect of open world games :)
To miss the joy is to miss all. Robert Louis Stevenson

User avatar
paramat
Developer
Posts: 3700
Joined: Sun Oct 28, 2012 00:05
GitHub: paramat
IRC: paramat
Location: UK

Re: Truly infinite worlds (poll)

by paramat » Post

I voted 'difficult to say' because both of 'actually infinite' and 'practically infinite' are both impossible. if something is 'practically infinite' it is infinite :)

Realistically, considering the work required, the way MT is and problems caused by larger co-ordinates, the only change with a reasonable chance of happening is to a s32 co-ordinate. Which means MC-size in all 3 axes, big enough, 4 billion nodes across (11 times larger than the orbit of the moon).

User avatar
rubenwardy
Moderator
Posts: 6972
Joined: Tue Jun 12, 2012 18:11
GitHub: rubenwardy
IRC: rubenwardy
In-game: rubenwardy
Location: Bristol, United Kingdom
Contact:

Re: Truly infinite worlds (poll)

by rubenwardy » Post

Minecraft manages practically infinite and runs better than Minetest (especially in bedrock)
Renewed Tab (my browser add-on) | Donate | Mods | Minetest Modding Book

Hello profile reader

User avatar
Stix
Member
Posts: 1385
Joined: Fri Aug 04, 2017 14:19
IRC: nil
In-game: Stix [+alts]
Location: USA

Re: Truly infinite worlds (poll)

by Stix » Post

I have an idea on how infinite worlds could (possibly) be generated. Lets say the world is made out of an infinite number of "mapblocks", lets further say these mapblocks are 32000^3. Pretend you walk across one mapblock into the next, once your far enough away from the first mapblock that no part of it is still rendered than the game unloads the whole mapblock.

One flaw with this is that if you leave a mapblock then return that area would be generated differently, right? Normally so, but what if the maps weren't random? What if each mapblock had its own "seed"? that way terrain would always generate the same way, and your system would only have to save the seed for each mapblock, just like if a friend told you the seed and cordinates of a nyan-cat, it would be there every time.

Now one other issue is that even if the terrain generates the same way when you load a mapblock again, your progress isn't saved, but theoretically this could be fixed by saving rendered areas you modified, and then overwriting the corresponding mapblock with it.

These are just my (stupid?) ideas on this subject, carry on.
Hey, what can i say? I'm the bad guy.

User avatar
Linuxdirk
Member
Posts: 3216
Joined: Wed Sep 17, 2014 11:21
In-game: Linuxdirk
Location: Germany
Contact:

Re: Truly infinite worlds (poll)

by Linuxdirk » Post

paramat wrote:[…] the only change with a reasonable chance of happening is to a s32 co-ordinate. Which means MC-size in all 3 axes, big enough, 4 billion nodes across (11 times larger than the orbit of the moon).
I’m looking forward to this since I first discovered Minetest. :)

User avatar
Hume2
Member
Posts: 709
Joined: Tue Jun 19, 2018 08:24
GitHub: Hume2
In-game: Hume2
Location: Czech Republic

Re: Truly infinite worlds (poll)

by Hume2 » Post

I have an idea how to store infinitely large coordinates. Let's take a Hilbert's hotel and place one computer into each room.

Then you could teleport to g_64 0 g_64 (where g_64 is the Graham number). Good luck with typing the Graham number in decimal :)

Please, change the title to "bigger worlds".
If you lack the reality, go on a trip or find a job.

User avatar
voxelproof
Member
Posts: 1087
Joined: Sat Aug 05, 2017 08:13
Location: Europe

Re: Truly infinite worlds (poll)

by voxelproof » Post

paramat wrote:I voted 'difficult to say' because both of 'actually infinite' and 'practically infinite' are both impossible. if something is 'practically infinite' it is infinite :)
When I was using the adjective 'infinite' without the qualification "practically" it was in order to add some flavour to this topic ;) However I'd argue that there's a sense to use the expression "practically infinite" as something separate from the idea of 'pure' mathematical infinity. So, let's introduce a definition of a "true practical infinity" in the open world games genre:

Code: Select all

An open world within a game is truly practically infinite when it's impossible for a player to see it whole or traverse by the means provided by in-game mechanics within his lifetime.
Thus this discussion contributed to the theory of gaming ;)
Realistically, considering the work required, the way MT is and problems caused by larger co-ordinates, the only change with a reasonable chance of happening is to a s32 co-ordinate. Which means MC-size in all 3 axes, big enough, 4 billion nodes across (11 times larger than the orbit of the moon).
By Voxel, it sounds really great. Good news :-)
To miss the joy is to miss all. Robert Louis Stevenson

User avatar
Dokimi
Member
Posts: 227
Joined: Mon Sep 19, 2016 03:52
GitHub: DokimiCU

Re: Truly infinite worlds (poll)

by Dokimi » Post

paramat wrote:.....the only change with a reasonable chance of happening is to a s32 co-ordinate. Which means MC-size in all 3 axes, big enough, 4 billion nodes across (11 times larger than the orbit of the moon).
Do we have a conclusion to the matter then?

If it is at least technically feasible, perhaps this should become part of the long-term roadmap? If we can we might as well? Someone will get to it one day!

Decision reached?

User avatar
voxelproof
Member
Posts: 1087
Joined: Sat Aug 05, 2017 08:13
Location: Europe

Re: Truly infinite worlds (poll)

by voxelproof » Post

Dokimi wrote:
paramat wrote:.....the only change with a reasonable chance of happening is to a s32 co-ordinate. Which means MC-size in all 3 axes, big enough, 4 billion nodes across (11 times larger than the orbit of the moon).
Do we have a conclusion to the matter then?

If it is at least technically feasible, perhaps this should become part of the long-term roadmap? If we can we might as well? Someone will get to it one day!

Decision reached?
I think this indeed should be a long-term goal, but the decision is up to the team. From my side I'd be happy to contribute my two cents by testing different mapgens and their variations. And the community may take into account my future screenshots, also in higher resolution and 3D :)
To miss the joy is to miss all. Robert Louis Stevenson

User avatar
Kilarin
Member
Posts: 894
Joined: Mon Mar 10, 2014 00:36
GitHub: Kilarin

Re: Truly infinite worlds (poll)

by Kilarin » Post

Previous related discussion: viewtopic.php?f=5&t=9183

User avatar
voxelproof
Member
Posts: 1087
Joined: Sat Aug 05, 2017 08:13
Location: Europe

Re: Truly infinite worlds (poll)

by voxelproof » Post

Dokimi wrote: Gameplay... I guess we wont know what a realistic Mt Everest map would be like until someone makes it! There is actually value simply in knowing that such mega-scale features are possible in a map. It adds to the ambience... the "wow" factor... even if only one nutter ever bothers to spend three wks climbing it! :-)
I've made a gaming experiment climbing mountains more than 1300 nodes high. Reaching the top usually took less than 20 minutes if the terrain wasn't too difficult. I think that the 8000 nodes high mountains is the most desirable size enabling to establish a new kind of gaming equally or even more engaging than most roguelikes (notice that the best roguelikes take months or even years of persistent gameplay to complete). Such climbing game with added hunger, thirst, weather conditions and exhaution would certainly provide great amount of fun and reaching the summit would give a player a strong sense of achievement and satisfaction. I don't think it's unlike roguelikes which are also a niche genre yet among them are the best games ever created.
To miss the joy is to miss all. Robert Louis Stevenson

User avatar
Dokimi
Member
Posts: 227
Joined: Mon Sep 19, 2016 03:52
GitHub: DokimiCU

Re: Truly infinite worlds (poll)

by Dokimi » Post

voxelproof wrote:
Dokimi wrote: Gameplay... I guess we wont know what a realistic Mt Everest map would be like until someone makes it! ..
... I think that the 8000 nodes high mountains is the most desirable size enabling to establish a new kind of gaming equally or even more engaging than most roguelikes ...
One of my early ideas for mg_tectonic was to have one giant mountain.

Maybe sometime I will make a simple version of this idea - if people are interested? Just to see what it is like. I likely wont take it much beyond the very basics, I'll leave it to others to build on the basic concept if they like it.

User avatar
voxelproof
Member
Posts: 1087
Joined: Sat Aug 05, 2017 08:13
Location: Europe

Re: Truly infinite worlds (poll)

by voxelproof » Post

Dokimi wrote:
voxelproof wrote:
Dokimi wrote: Gameplay... I guess we wont know what a realistic Mt Everest map would be like until someone makes it! ..
... I think that the 8000 nodes high mountains is the most desirable size enabling to establish a new kind of gaming equally or even more engaging than most roguelikes ...
One of my early ideas for mg_tectonic was to have one giant mountain.

Maybe sometime I will make a simple version of this idea - if people are interested? Just to see what it is like. I likely wont take it much beyond the very basics, I'll leave it to others to build on the basic concept if they like it.
Go on! :)
To miss the joy is to miss all. Robert Louis Stevenson

User avatar
Linuxdirk
Member
Posts: 3216
Joined: Wed Sep 17, 2014 11:21
In-game: Linuxdirk
Location: Germany
Contact:

Re: Truly infinite worlds (poll)

by Linuxdirk » Post

Dokimi wrote:One of my early ideas for mg_tectonic was to have one giant mountain.
We have roughly half of the maximum height over ground. It's a pity that the highest "mountains" are only a couple of hundred meters high while having 32 kilometers available height just being empty space.

User avatar
Dokimi
Member
Posts: 227
Joined: Mon Sep 19, 2016 03:52
GitHub: DokimiCU

Re: Truly infinite worlds (poll)

by Dokimi » Post

voxelproof wrote:
Dokimi wrote:
One of my early ideas for mg_tectonic was to have one giant mountain.

Maybe sometime I will make a simple version of this idea - if people are interested? Just to see what it is like. I likely wont take it much beyond the very basics, I'll leave it to others to build on the basic concept if they like it.
Go on! :)
I realized that mg_tectonic could be adapted for this purpose. It should be straightforward... (should!). However, this is not a priority at the moment. I will post how to do this on the mg_tectonic thread, and leave the challenge to someone else.

User avatar
aristotle
Member
Posts: 79
Joined: Wed Mar 14, 2018 23:40
GitHub: askotos
IRC: aristotle_
In-game: aristotle
Location: Currently on Melpomene, waiting for the starship to be fixed.
Contact:

Re: Truly infinite worlds (poll)

by aristotle » Post

rubenwardy wrote:Minecraft manages practically infinite and runs better than Minetest (especially in bedrock)
Since when? :O
I knew they could not build higher than 256 and lower then -256 at least up to v1.12...
Horizontally does not mean that MineTest allows less: having a y coord in the range[-256, 256] is quite limiting. Going up and down so much in the range of thousands is definitely a great plus MineTest has (or at least has had for quite a while) over MineCraft.

I definitely enjoyed the speed boost of MineTest on my old i3 notebook when comparing it to MineCraft which did not use memory so well (the Java virtual machine is a beast by itself). Thus, though I have no official benchmark , my creative experience in both the games has been definitely different and... with no doubt more pleasurable in MT!

EDIT:
Though I have been building for a few months underground I have never reached the lower limit.
In MineCraft the bedrock layer is practically a necessity: it is in fact annoying to start falling every time you go down a bit more... Making you noticing you are about to fall into the void is compulsory there. We, on the other side, have lots of underground stuff: caves, basins, ruins, ... and lots of other stuff on surface such as floating isles, the asteroids belt, ....

BTW This does not mean I would not like having world size biomes and a way to join a couple of them together like if they were 2 or more dimensions (maybe with a different number of suns and moons.

EDIT2:
Probably the MT map generators have not yet exploited so much the potentiality of all this MT verticality, thus underlining the smaller horizontal size.
I have not tried so many mapgens, nor customized their parameters so much, ... but they generally provide environments that tend to stay in the MC elevation ranges, probably because of the typical users expectations, but I think they are still to show what they might be capable of if they just started extending their ranges (caverealms* and other_worlds are already trying/doing it): lakes, falls, rivers, snow, ice and desert depending on more general rules of altitude and longitude might immediately give us a whole new set of biomes...

We can go higher than Himalaya and lower than 10.000 meters below the sea level...
We might have cool underwater worlds if just some water rendering problem were solved - i.e. concerning sea plants, slopes rails, ... -
Solving this problem might improve MT capabilities as much as the circular/table saw has done, and probably more from the user's point of view.
The circular saw requires the end user / player to learn how to use it, a full working underwater world just asks to be explored and leaves its implementation details to skilled modders/developers and for this reason more quickly unleashes its full potential to the community.

For this reason I am not so sure the map should be infinite, probably just a bit larger - in the range of the 4.000 km as it has already been suggested. I would be more into solving the water problem first: because of that a full set of biomes is missing from MT, wasting most of its [vertical] potentiality.
Happy builds & explorations! | Initiating my Creative Minetest channel on YouTube

Nutty8
Member
Posts: 10
Joined: Wed Nov 08, 2017 16:08
In-game: Nutty8
Location: Brazil

Re: Truly infinite worlds (poll)

by Nutty8 » Post

Krock wrote:Hume2, this idea was already mentioned in various topics but is finally not usable because
1) Objects can stand in between nodes, so we would need likely a 128 bit floating point number for such big maps (takes much computing time)
2) Breaks the map format. It would need a conversion from old to new (takes a very long time on large maps)
3) The potential of garbage increases enormously: Admins teleport to a far away location and the mapblock will never be used again
4) Effort to clean up above mentioned, unused mapblocks
5) Very painful to teleport far away. Where was my house again? 932355235,24,-53263443 or 932356235,24,-53263443 ?
6) Our code base is not made for this. It would mean rewriting basically all code with position values and Irrlicht itself if such enormous sizes are wanted.

Some server admins already limit their map to 5 or 10 km to limit the map size (garbage mapblocks) and to keep the players together. Already now it's painful without teleporters to get anywhere distant if there's no railway or street nearby.
So please, please could we just drop this idea? 64 km ought to be enough for everyone.
I voted for infinite maps, but after reading this, everything makes sense now.
Why a map so enormous? I didn't explore even 10% of my maps :')
In love with Minetest

User avatar
Wuzzy
Member
Posts: 4786
Joined: Mon Sep 24, 2012 15:01
GitHub: Wuzzy2
IRC: Wuzzy
In-game: Wuzzy
Contact:

Re: Truly infinite worlds (poll)

by Wuzzy » Post

I bet anyone voting for infinite probably hasn't even explored 5% of the map or ever legit ran out of space because the map was too small.

Besides, what's the point of an infinite world if there's not much to do inside?

A sandbox does not become better because it's larger, it becomes better by adding more things to do in it. The size of a sandbox isn't that important. What is inside the sandbox is way more important to the gameplay.

https://www.youtube.com/watch?v=efKxC0RgyV4

I claim even if Minetest's world would actually be infinite (which is impossible) it wouldn't be a real game-changer. The gameplay would still be largely the same. Think about it: What if the world would be infinite? How would it affect gameplay? I think not very much. Probably not at all.

60000³ isn't small, it takes a crazy amount of time to just explore just the surface area and there's usually not a real reason to do so. Even just walking from one end to the other takes a long time and a lot of dedication, so usually you don't do it.

Perhaps we should just lie to the players by saying Minetest's worlds are truly infinite and hope nobody figures out. ;-) There's a good chance many players who just play won't notice. XD

User avatar
Linuxdirk
Member
Posts: 3216
Joined: Wed Sep 17, 2014 11:21
In-game: Linuxdirk
Location: Germany
Contact:

Re: Truly infinite worlds (poll)

by Linuxdirk » Post

Wuzzy wrote:(which is impossible)
Actually it is not impossible to create a generation system that allows creating and addressing finitely large worlds. It is all a question of technical practicability.

Switching to a 64 bits addressing system would be a first step in the right direction but “muh backwards compatibility” (it would be best working on an actual 1.0 version after releasing 0.5 under the bogus version number 5.0 that actually uses a 64 bits addressing system).
Wuzzy wrote:What if the world would be infinite? How would it affect gameplay? I think not very much. Probably not at all.
Yeah, or imagine a space exploration game having 18 quintillion different planets. It would take the whole community working together 24/7 a few billion years to explore all of said planets – ridiculous! … but still one of the main selling points of No Man’s Sky (that recovered very well after the miserable launch back in 2016).

It’s not about needing unlimited (or absurdly large 64 bit addressed) worlds. It’s about having the option to have them.
Wuzzy wrote:it takes a crazy amount of time to just explore just the surface area and there's usually not a real reason to do so.
Yeah, and that is sad. Even with the most recent and advanced world generators it’s just repetitive and boring. Yes, there are some awesome structures underground and overground, but be honest: Below -1024 and above ~+1000 there is nothing new to see, or literally nothing other than air. And after seeing all overground biomes for one or two times there’s nothing new either.

So we have a large 64k³ sandbox that is only interesting within a ~2000 nodes radius around 0,0,0.

User avatar
voxelproof
Member
Posts: 1087
Joined: Sat Aug 05, 2017 08:13
Location: Europe

Re: Truly infinite worlds (poll)

by voxelproof » Post

Linuxdirk wrote:
Wuzzy wrote:it takes a crazy amount of time to just explore just the surface area and there's usually not a real reason to do so.
Yeah, and that is sad. Even with the most recent and advanced world generators it’s just repetitive and boring. Yes, there are some awesome structures underground and overground, but be honest: Below -1024 and above ~+1000 there is nothing new to see, or literally nothing other than air. And after seeing all overground biomes for one or two times there’s nothing new either.

So we have a large 64k³ sandbox that is only interesting within a ~2000 nodes radius around 0,0,0.
But this doesn't have to be true. Look at the "Heaven" tweaking of Gael's valleys mg: it generates archipelagos of sparsely scattered interesting and explorable islands and it's obvious to me that for this set of parameters the present size of MT worlds is too small. And imagine a mapgen which creates real-sized huge rivers hundreds of thousands nodes long, huge deserts and so on.

You would be right if the generated worlds would look everywhere alike. But the idea of exploration is not that the terrain must be interesting everywhere: the whole joy of discovery lies in finding those 1% of really awesome spots which make exploration rewarding. I am still convinced that larger worlds (like those in Minecraft in terms of horizontal dimensions) should be given a good try.
To miss the joy is to miss all. Robert Louis Stevenson

User avatar
Insurrection
New member
Posts: 6
Joined: Sun Mar 05, 2017 02:13
In-game: Insurrection

Re: Truly infinite worlds (poll)

by Insurrection » Post

"Truly" infinite is "truly" impossible because it requires "truly" infinite data storage, etc. As brought up before, though, there is "practically infinite". However, I don't think it even needs to have Minecraft's horizontal size - I'd be content with anywhere from 2x, 4x, or 8x being simple options for world generation, especially for servers that wouldn't mind the increase in storage load. As said before, the world is already pretty huge, and it'd take a significant amount of time to go from one side of the world to the other (or even the center to one side).

User avatar
Linuxdirk
Member
Posts: 3216
Joined: Wed Sep 17, 2014 11:21
In-game: Linuxdirk
Location: Germany
Contact:

Re: Truly infinite worlds (poll)

by Linuxdirk » Post

Insurrection wrote:I don't think it even needs to have Minecraft's horizontal size
But why not? Using a proper addressing system as described earlier it would not matter if you want to walk 2 or 2 billion nodes in one direction. If the system would ever be rewritten we should aim for the maximum possible size. And if that be 10 times the width and length of a Minecraft world (and this value also for the height) I’d say DO IT. It’s all about to have, not about to need.

User avatar
Insurrection
New member
Posts: 6
Joined: Sun Mar 05, 2017 02:13
In-game: Insurrection

Re: Truly infinite worlds (poll)

by Insurrection » Post

Yeah, true. Definitely would be nice to have a world that has the same surface area of a gas giant, if only for the "because we can" reason. Still would be up to mod/game makers to make that space interesting, though, but I feel plenty of people (myself included) are confident enough to be up to the task (:

Post Reply

Who is online

Users browsing this forum: No registered users and 6 guests