Mapgen mods rules

[bell07]

This issue is to speak about applying mapgen mods in general to the whynot game.

We need a strategy for Rule 1 with mods tracked in #32.

I have different worlds with Whynot, that I see as "Base game", with additional mods enabled.
In detail:

• 3x Whynot Base worlds with different mapgens
• 3x Unternull/Startanode worlds with water, lava and air ground
• 1x Ethereal world

If we merge ethereal mod for examlpe, I get 6 of 7 worlds broken.

My proposal is to create subgames for different mapgen enhancements. Maming proposal:

• whynot-game (as we have)
• whynot-void-game
• whynot-ethereal-game
• whynot-underch-game
• whynot-luscious-game

The subgames could be whynot fork, or my preferred: whynot-game is a mod source for the subgames.
All subgames can be placed innto minetest-whynot organization at github.

Maybe the solution for #81 is to setup the whynot-underch-game?

@Lazerbeak12345, @dacmot , what do you think?

I started already to implement whynot-void-game some times ago, but still unfinished, need to clearup and update my unternull fork at the first.

[dacmot]

To be honest, I'm not crazy about the idea of trying to maintain many different games. On the other hand, I agree with you that adding mapgens will possibly break previously generated maps.

Could we have only two? The standard whynot-game, as it is now, without any mapgens. And another we could name something like whynot-game-plus with all the under/above ground realm ideas we want to cram in there.

[bell07]

Any subgame need a responsible maintainer in team that is motivated to maintain it. No new game should appear just to close "Add mapgen mod xy" issue.

I personally like to maintain the whynot-void-game game, but cannot provide ETA if the game appears. Maybe I should rename it to whynot-game-minus ;-)

If you like to maintain the whynot-game-plus, then the game can appear. Which mapgen mods do you see in plus?

I do not think a whynot based subgame need much time for maintenance, since the most mods are from the same source.
The reason why "void" takes longer is because I plan to sort out my unternull fork at the first, since lot of changes was appear upstream.

For example, current state of the "void" game's build-mods.sh

#!/bin/bash

PROJ="$(realpath $(dirname $0)/..)"   # Absolute path
SRC="$PROJ"/builder/mods_src/
DST="$PROJ"/mods/            # Subgame mods
LOG="$PROJ"/mod_sources.txt

source "$PROJ"/builder/mods_src/whynot-game/builder/build-whynot.lib
cd "$SRC" # for proper resolving the '*'

# Remove old log
rm "$LOG" 2>/dev/null

## Sync whynot_game
mod_install ambience      whynot-game/mods/ambience/*
mod_install decor         whynot-game/mods/decor/*
mod_install flora_ores    whynot-game/mods/flora_ores/*
mod_install food          whynot-game/mods/food/*
mod_install libs          whynot-game/mods/libs/*
mod_install mapgen --exclude cloudlands whynot-game/mods/mapgen/*
mod_install mesecons      whynot-game/mods/mesecons/*
mod_install minetest_game whynot-game/mods/minetest_game/*
mod_install mobs_redo     whynot-game/mods/mobs_redo/*
mod_install player        whynot-game/mods/player/*
mod_install tools         whynot-game/mods/tools/*

## Void mods
mod_install void startanode

## Renewable mods to get items in survival
mod_install renewable frozenenergy
mod_install renewable techpack/gravelsieve
mod_install renewable unternull/crushingfurnace

1 mod excluded, 4 additional mods provided.

With .gitmodules

[submodule "builder/mods_src/whynot-game"]
	path = builder/mods_src/whynot-game
	url = https://github.com/minetest-whynot/whynot-game/
[submodule "builder/mods_src/startanode"]
	path = builder/mods_src/startanode
	url = https://github.com/minetest-mods/startanode
[submodule "builder/mods_src/frozenenergy"]
	path = builder/mods_src/frozenenergy
	url = https://github.com/minetest-mods/frozenenergy
[submodule "builder/mods_src/techpack"]
	path = builder/mods_src/techpack
	url = https://github.com/joe7575/techpack
[submodule "builder/mods_src/unternull"]
	path = builder/mods_src/unternull
	url = https://github.com/CasimirKaPazi/unternull

And of course own minetest.confsettings;-)

[Lazerbeak12345]

I don't think we need to make a new game for each of these mapgen mods, as users can do exactly what you've done, and just add the mod to a whynot world.

As for whynot-void or whatever we end up calling it, it sounds like a lot of work.

[dacmot]

If you don't want to retire quite yet, and create various games based on WhyNot, I'm not going to stop you ;-)

For the mapgens, I was thinking plantlife and luscious for the surface, caverealm_lite (-1500, 20000) and nether (20000, 32000) for the underground. I'd like to include underch as well, pending bug fixes and the ability to specify depth limits. Probably squeeze it between caverealm and nether.

Since the realms mod was never in a usuable state, I was wondering if it would be possible to make the cloudlands/floating islands with biomes from etherreal, without etherreal being on the ground.

[bell07]

I just try to keep existing world playeable, even I do not play them ;-) The void game is not from scratch, is quite old, but was not published in the past. I decided to revise, update and tidy my unternull fork before I can publish the game.

In most cases it is enough just enable some additional mods to get new world flair. No own subgame required in this case. This way I handle my ethereal world for example.

If mods needs to be removed from the game because of incompatibility, the only way I see is to create new subgame. In case of "void" I exclude cloudlands for example and set other defaults in minetest.conf.

If you add plantlife and luscious for the surface to whynot, I need to create new (whynot-classic?) game that excludes this mods, to get the old MTG-like flair back. I prefer not to go this way, but it is up to you, since I retire for whynot

[Lazerbeak12345]

From recent testing on adding things like caverealms, plantlife, nether and much, much more to a live server with actual players; this is what I've learned.

• Adding new mapgen mods didn't effect the block content of old chunks. This is both a bad and a good thing. Good because the world isn't destroyed, bad because now you get oddities like some mobs don't spawn in the correct locations, or portals taking you to chunks that don't have the expected biome.

I even tried tweaking things like layer depth for the nether. What happened is any chunks generated before that change didn't spawn some nether mobs (I had three mods that added nether mobs, balrog, "nether mobs", and the "whilhimes nether" thing.) anymore, so those areas became boring.

• Mobs are strange. My world got partially destroyed by low-quality spawning logic. Also, as mentioned before, mobs didn't always spawn where I wanted them to.

• Some mapgen mods break others. I looked a lot but I couldn't find any plantlife plants when ethereal was present.

• Some mapgen mods are intended to be played in a specific order. When the sky island mod (whatever it's called) is present, the nether mod provides portals to teleport a player to that location from the nether. If we add nether we will need to raise the y-level of those islands, since as it is, players can easily build up to where they are, so they aren't really late-game.

• Still missing something. Despite all of these mods I added, none of them really improved exploration, aside from the nether mod and ethereal. The nether mod adds plenty of unique hard-to-find things that can be used for future gameplay progression. Ethereal adds these as well. The rest of these mods improved dungeons, or really added any structures at all. There wasn't anything driving exploration on the surface, where most mapgen effected things, so most players didn't notice the new things as they were added. The new mapgen features were all around pretty boring, aside from nether (and perhaps caverealms?). Gives you places to go, but once you are there, there's nothing to do.