Thanks phr00t 
As this thread started to live again, I want to give a little status update:
First I glued textures onto my blocks.
After having a little break, I implemented the Entity System basics this weekend. As tomorrow is a free day in Germany I hope to fill it with live in a way that one can eventually walk onto the terrain. 8)
@nihal said:
I have been wanting to use a block terrain for my game as well. But seeing that one half of jMonkeyEngine devs are working on block terrains and the other half is working on space games, I am trying hard to come up with something a bit more alternative...
The huge thing not to forget is that "Block" and "Space" games are only sub-genres or you could call it "setting". You could compare 2 space games have 2 different things, the same goes with block games. Look at Mythruna and 3079. Both block games, 2 different games.
So, in the end, it's a matter of implementation, story and game play that will make your game stand out, not its settings or the type of environment its engine will generate. Of course the latter could complement the former if done right.
Here's a thought for you, why not make a blocky space game? ;)
@madjack said:
Here's a thought for you, why not make a blocky space game? ;)
Heheh... after all, that's what Mojang is doing next. :)
Beat them to the punch then. I have seen screenies of that…thing and… not convinced. Not that I’d be interested in whatever they might publish anyhow.
It’s more Notch doing stuff that he wants to do than Mojang officially. But of course he will use his company to publish the game once it’s finished
I like the ideas. I’m intereseted in what it will evolve.
@madjack said:
Beat them to the punch then. I have seen screenies of that...thing and... not convinced. Not that I'd be interested in whatever they might publish anyhow.
People have made space ships in Mythruna that are stunning. The slopes and stuff really play well there.
@pspeed said:
People have made space ships in Mythruna that are stunning. The slopes and stuff really play well there.
I don't doubt that. :D
After a long time and a lot of exams I reached a step where I feel confident with my work again:
I added the EntitySystem, Physics and the synchronization of the server entites with the client including lag prevention methods. First it performed really bad but when I put generating the MeshCollisionShape into the asynchonous generation thread it became very nice 
Screens are not very helpful when showing physics, but nevertheless I include one^^
http://arkaryx.net/screens/physicsWorking.png
The next steps are now integrating the player’s model with it’s animations (later I plan to include limb exact collision detection) and of course vegetation and zivilisation in an improvement generator.
BTW: VisualVM has proven to be a very useful tool when detecting performance bottlenecks.
Great! Can we expect a prototype video submission?
When I get the animated character walking soon, I’m tempted
hello, this is an interesting project. I am also trying to make an random content rpg.
Here are some things i noticed:
procedural generated quests.
With procedural generated quests you can create the “basic gameplay” of them, e.g kill 40 rats
but they will lack :
- story: it has to created by hand, else it will be the similar.
- lore:
a) what game npc can participate in this quest.
b) repercussions that these quests cause to them / to the world.
these have to be written by hand / players.
Yes, you are right. On my website I specified it a bit more precisely: “Half-procedural generated quests. Not just grinding like one would expect!”.
I’m aware that one needs the creativity of human beings with emphasis on the plural to create an interesting mix of tasks and story elements. So the idea is to take basic elements, combining them and offering them in more or less random order. Similar it should be with the story.
I’m curious how that will evolve myself^^ The most technical stuff is done now, so that the game logic part comes closer eventually.
Actually there’s no reason entirely random quests couldn’t be constructed by the computer. Think about how the complexity of drawing a modern 3d scene would have sounded to someone 20 years ago. The quest problem is far less complex than that - it’s just had much less time and effort spent on it as the return on investment is less clear.
Look at things like the left 4 dead Director for examples of where things are heading, or @phr00t 's Gentrieve 2 which is essentially doing procedurally generated quests.
The Left 4 Dead Director is really cool. It inspired me when I read about it. And it’s cool how they tested it, with electric measurements on players so they always have kind of a sinus curve of tension.
entirely random quests
There would always be an underlying compute pattern that dictates it,
lets say you spend 1 week to create that pattern, good after 5-20 quests the pattern
can be recognised by the player and quests become boring.
Thats why quest story must be written by humans.
left 4 dead Director
It doesn't do some unique, its a series of heuristic functions that modify the game difficulty.
The gamut of quests is limited, that they are done by humans or not. That doesn’t matter at all.
What makes a difference, or could actually, is the storytelling that human beings possess. With that said, it’s still possible to make good random quests. The thing is, lots of time will be needed to come up with a scheme that is working and not too repetitive.
I’ve started working on that several months ago. It’s mostly fetch quests for now, but the goal was to install the groundwork first so I could verify the correct functioning of the quest manager. It is working.
The next step will be to implement a (very) wide array of quest types and tie those with random elements and NPCs so it’s not just “Go fetch this” or “Go kill that”. The ultimate goal will come from an NPC (or group of) need. Depending on the results of the draw the randomizer could attach secondary or ternary steps.
But that’s a long way away and as with anything else, it’s as good as the time you spend on it.