Stranded III Dev. Blog - Comments
Offline
Forum
Projects Stranded III Dev. Blog - CommentsI won't implement the urge to go to toilet periodically in Stranded III because I think it isn't a fun mechanic. There will be related mechanics though. Even Stranded II already had the mechanic that poo spawned some time after you ate a plum.
edited 1×, last 15.05.23 10:02:58 pm
DC: I agree that it is not a fun mechanic. It's already inconvenient enough irl, so to T-69 and Mora: It would not add value to the game.However, some system of animal manure for plant fertilizer may be handy. We mainly use that irl, and for practical reasons: animals make much more of it than humans.
Also,
DC, about the AI idea I had: What kinds of AI-supporting platforms, if any, are there that could be be used to import some simple ML script or programs, for training, reinforcement, etc? I'm thinking of the possibility of pre-training animal behavior before release, and/or during runtime (I am unaware the latter is feasible, though). Of course, there is also the artificial life approach (https://en.wikipedia.org/wiki/Artificial_life), which is appropriate given that, for the behavior of animals, you are trying to create something that resembles life. There are many links within that article that may contain applicable knowledge as well. And I found ALife resources here: https://alife.org/online-resources/. That may come in handy if you consider that approach. ModJuicer: Stranded III "AI" will use simple state machines. There will definitely be no machine learning or neural networks or whatever involved. The reasons for that? I prefer having full control and understanding as well as implementing logic myself. I also think that ML is quite wasteful because it costs much processing power to train AI and probably also more memory and CPU to run it than my approach. Furthermore I can't afford to look into that topic enough - it would cost too much time.It will probably be possible to script AI with Lua though. I'm not sure but it might be possible to create a machine learning interface that way if you really want to.
For that reason, I drew a concept menu to provide an example of what could be implemented with the states/effects in S3.
The concept menu allows effects of any size, any color, any place relative to the entity that carries the state, any audio file, and the ability to adjust volume. This example allows the sprite itself to be changed so much that, if it was implemented, there may as well also be the ability to create new, custom states/effects from scratch.
Now, it looks fairly cluttered (because of my wonderful artistic ability
), but it would look neater in the actual game. Anyway, this is intended to represent an improved version of the menu that shows up when you right click on an effect applied to a given entity in the editor mode of Stranded II, like that which I would expect to see in Stranded III. Only relevant settings would be displayed (i.e. if effect does not create sprites/particles then color, size, offset, sprite, and sprite scale options would be omitted. If speed is the option rather than strength, that is the option that would be displayed instead, and so on), and it would allow much more customization and freedom. Of course, there are a lot of modifications that can be made to this menu if you wish, as this is just one possible form, but this is the general idea. You could even extend on the idea and make effects act dynamically so that they interact with their environment. I.e. a fire shrinks slowly over time as fuel dwindles, from a pillar of flame down to a modest glow from within the coals, among other possibilities. Maybe even things like what CS2 did with volumetric clouds if you're feeling ambitious.Does this idea seem viable and/or preferable to you?
EDIT2: I had an idea of a feature that could be indispensable to unexperienced coders on Stranded III: If you made so, in a given setting or menu in editor, it would auto-generate code that would have the same exact effect as a the setting being set.
For example, if you were setting a color, there would also be a button you could press that would spit out the code equivalent of setting the specified color, so it can be integrated into any code in the future. You could implement this capability into basically any given setting or menu, so that the desired code equivalent would be just a copy-paste away.
It would help people learn the scripting language with ease, without having to even go through a tutorial or anything of the sort (and I do recommend including a scripting tutorial within the game, especially if you are planning on adding new code features or additions which don't already exist in LUA). It would also just be convenient, and could serve to assuage the worries and fears of perfectionists.
Also, some things just cannot easily be coded manually, and are virtually impossible to get right. Being able to get it right the first time for those types of things would be vital to automating tasks. It would also open up opportunities to integrate more things into the code itself, and thus improve the functionality and of it, since complexity or difficulty would be less of a barrier in the way of integrating such auto-generated code.
Alongside this, features can be added which more intuitively augment specific tasks being performed by code, which output the resulting code. This could be useful for camera sequences, custom entity behavior, and well-integrated mapmaking.
The end result would be a script that can do more, is easier to use, and is far more accessible to people. People would be far more likely to start scripting to begin with in such a game environment.
I can personally assess to that. At five years old, around the time I had first played Stranded II, I had tried for hours to get the code to do something (I did not actually know how it worked. I just typed in things like "give me unlimited health" and "make the parrots slower", which, looking back with my full knowledge of s2script, was pretty cute), and never really got anywhere (I finally learned how s2script works at ~12 years old). However, given the same circumstances, but with auto-generated code and intuitive features to auto-generate code, I most certainly would have learned how to script at five years old. I hope Stranded III will be able to be what I lacked back then.
edited 14×, last 21.03.24 02:37:08 am
No multiplayer? I think thats okay, modding scene might do it if to source will be open. I was wondering why there is no update for a looong time, I guess learning Godot was time consuming. I never liked Unity as it was, thats why I was messing with NeoAxis and tried to "port" S2 to it but i dont have time to learn it. Good luck and I am looking for future updates! Cheers!