Is Desolation Open Source? I want to help on the project.

[Luna.]

I've been enjoying the game this week, and although I enjoy both reborn and desolation, I feel like desolation's story is better. Playing desolation, it feels like I'm almost playing a new game, that's better than pokemon (although I havent played much bc vanilla pokemon is boring and too easy). That being said, I wouldn't mind learning how to contribute to this project when I have time. I enjoyed it, so I don't mind giving back, and also help to develop something I liked. I do have programming skills but not very familiar with rpgmaker, but I could pick it up fast.

 

I'm close to graduating and getting my degree in computer science, so I'm confident in being able to pick up the necessary info/skills to invest into this project.

 

Things I'm interested on working on:

1. custom trainer ai - different levels of trainers, unique trainers having unique ai to leverage their teams effectively. (switching mind games etc, making you feel like you were fighting an expert human player). I think this would add a lot of replayability to the game and make it much much more interesting.
2. better breeding - this feature would be added on top of the normal breeding system.

• I think you should be able to invest more into the pokemon daycare, so that eggs are automatically deposited into boxes. (can set which box to deposit eggs in). Can also balance it by forcing you to pay credits for each egg deposited as well.
• Investing credits so that you can hire a hatcher who hatches eggs for you. (Up to a certain amount a time) - using the ev training room in the manor. You pay in cash after hiring him. by default he breeds from one box, but you can spend an exorbitant amount of credits so he can breed from another box as well. (selecting the boxes you want to breed from).
• You can upgrade the hatcher so that he can hatch eggs faster for you too/give him a pokemon with an egg hatching ability.

1. optimization and qol fixes - trying to optimize performance in areas such as celia and blackview if possible after the first 2 are done. I have a $1200 pc and I notice frame drops. And fixing minor annoyances if I notice any. ie. making repels last longer.
2. psychologist - could be way more expensive, but you could choose the nature you want. make it less grindy, but also make it so the price would be so that you would have to think whether it's worth it or not to do it.
3. battle universities - elite places of study where you can pay a large amount of money to enter and learn about certain battle tactics about that schools unique Pokémon + the opportunity to earn them - competitions between schools for prizes/benefits. You can only join one. Probably located in celia/other places. Access to certain tms/move trainers (they can share some in common, but each school can have unique pokemon and tms that are very desirable)

Although these changes would make the game less grindy, I think the difficulty would still be the same, just saves time. And I think the difficulty would even increase with trainer ai upgrades

 

[Yumil]

Desolation isn't open source. That being said, Caz seems pretty open to external contributions from my exchanges with him, so I'd say it's definitely worth trying to get in contact with him and see if he'd approve of those features.

 

If you want to get started on working, despite deso not being open source, the only thing you need to get working is RPG maker XP. if you have it, you can just drop a project file in the game folder and open it and you'll be able to edit the project freely. The only thing you'd need that isn't included in the game folder are the PBS files handling things like the trainers teams and so on, but those are available to download somewhere in this forum, so that's not a problem either. From there on, your only limit is your knowledge of ruby, Rpg maker's limitations, and your patience.

 

Two piece of advice if you feel like going down this path :

A) When coding new features, I heavily suggests making them as detached as possible from the current scritpts - since you'll be working on your own version of the scripts, someone is gonna have to port them over to deso's main version at some point, so the more your code is organized in a way that lets you do that easily, the better.

 

a personnal habit of  mine whenever I had to edit existing scripts (since sometimes you don't have the choice, after all), is to bracket every modifications between lines of ### where you note your name/handle, the feature this is supposed to help with, and a unique number for each modification. It makes it way easier to retrieve those modifications later on with a CTRL+SHIFT+F, and coupled with a documentation file where you list all those modifications and explain a bit what they're supposed to do, make it way easier to port code.

 

B)Regarding the AI, if you really intend to go down that way, you might want to get in touch with Andracass from Reborn as well, since she's done a lot of work with the AI already and Deso uses Reborn's AI code and there's a lot of modifications she's passed to the Deso team since E5's release that aren't implemented in E5. the Battle AI in general is a pretty tough cookie to crack, from my understanding, so I'd just advise to wait to gain some understanding of the general structure of the scripts to try anything substantial with it. 

 

Those features sounds like some interesting ideas, honestly.I really like the better breeding one.

[Luna.]

11 hours ago, Yumil said:

Desolation isn't open source. That being said, Caz seems pretty open to external contributions from my exchanges with him, so I'd say it's definitely worth trying to get in contact with him and see if he'd approve of those features.

 

If you want to get started on working, despite deso not being open source, the only thing you need to get working is RPG maker XP. if you have it, you can just drop a project file in the game folder and open it and you'll be able to edit the project freely. The only thing you'd need that isn't included in the game folder are the PBS files handling things like the trainers teams and so on, but those are available to download somewhere in this forum, so that's not a problem either. From there on, your only limit is your knowledge of ruby, Rpg maker's limitations, and your patience.

 

Two piece of advice if you feel like going down this path :

A) When coding new features, I heavily suggests making them as detached as possible from the current scritpts - since you'll be working on your own version of the scripts, someone is gonna have to port them over to deso's main version at some point, so the more your code is organized in a way that lets you do that easily, the better.

 

a personnal habit of  mine whenever I had to edit existing scripts (since sometimes you don't have the choice, after all), is to bracket every modifications between lines of ### where you note your name/handle, the feature this is supposed to help with, and a unique number for each modification. It makes it way easier to retrieve those modifications later on with a CTRL+SHIFT+F, and coupled with a documentation file where you list all those modifications and explain a bit what they're supposed to do, make it way easier to port code.

 

B)Regarding the AI, if you really intend to go down that way, you might want to get in touch with Andracass from Reborn as well, since she's done a lot of work with the AI already and Deso uses Reborn's AI code and there's a lot of modifications she's passed to the Deso team since E5's release that aren't implemented in E5. the Battle AI in general is a pretty tough cookie to crack, from my understanding, so I'd just advise to wait to gain some understanding of the general structure of the scripts to try anything substantial with it. 

 

Those features sounds like some interesting ideas, honestly.I really like the better breeding one.

I can go grab the necesary tools. and well I do have some experience working with machine learning and neural nets, but I dunno if that's the approach here. Would prob just try writing an algorithm. Looks like pple at standford etc have already been tinkering with stuff:

https://web.stanford.edu/class/aa228/reports/2018/final151.pdf
https://www.smogon.com/forums/threads/technical-machine-a-pokemon-ai.69021/
https://www.geeksforgeeks.org/minimax-algorithm-in-game-theory-set-3-tic-tac-toe-ai-finding-optimal-move/

 

It said their bot managed to reach 1300+ elo on pokemon showdown, I dunno how good that is bc I havent battled competitively but if its beating players a bit that's good.

For an expert trainer level, the bot might be able to calculate dmg from you and the pokemon, and then decide if it wants to switch or not, or have a gameplan. And the good ones can try to guess what pokemon you're trying to switch into by looking at your team that you've showed so far, and use a move/set up that will hit it. Nasty.

 

That kind of ai would prob be only for gym leaders/main characters. The average people can have an ai that makes mistakes/plays like average/noob.