Ren'Py My World Isekai [v0.0.791] [Silac]

[dgiriyftes]

Yeah, basically haha. For now it is atleast. I'll add some "trait personality" to make it less random.
I'll try to get the 0.1.0 as soon as possible!

Well, it's not that random.
The NPCs have 3 interactions possible between eachother.
Talking, insulting and Fighting.
When two NPCs do not know each other, they have a 50% of chance to engage a positive or negative interaction.
But then, the odds depends on their relationships/affinity. The more friendly an NPC is toward another, the higher chance he will start a positive interaction

I might have a suggestion for you. But of course do whatever you want. It also might very well not be compatible with your current setup or might be to big of a change, but I feel like it could serve as inspiration. I am also not sure how well it would work performance wise.

I feel like the aspect of randomness can give replay value. I had a brief look at your game and all the NPCs governed by some discrete states e.g. where they are and what they are doing. A possible idea to make the characters distinct while retaining randomness might be the following.
Say a configuration of the system consists of the locations and lvls of all characters and what they are doing. On top of this each of your characters might have internal values describing their personality.

Than one might be able set up Markov chain Monte Carlo scheme to transition from one configuration to the next. I assume you have some "energy function" being the sum of how well each character is doing (the quality of your city could be included in this as well, as I saw in your devlog that you wanted the city to be variable as well). This energy function could also be higher or lower based on if characters "like" to be in this situation due their personality. So this would assign an energy value to each configuration. Then a transition rate from one configuration to the next could be described by the Boltzmann factor of your difference in energy between the states, as in the Metroplis-Hasting algorithm. (So a change from your current configuration to a randomly chosen different configuration would always be accepted if the energy decreases but if it increases it might still happen based on the Boltmann factor.)
So this transition would be accepted or rejected based on some random number which is higher or lower than its correspond transtion rate.
(More explantion at the Metroplis-Hasting algorithm wikipedia page)
Of course most states should be unreachable from any current one, so one could optimize here. And many decisions of NPCs are independent so these could be performed in parallel.

So this would minimize the energy function while admitting fluctuations due to randomness. I'd assume their would be a lot of metastability, so one could end up with "stationary" configurations which look quite different from one another based on the chosen seeds for the RNG.

 

[Silac]

Cute art, quite motivating even without much content to speak of. Reminds me maybe a liiitle bit of Unholy Academy.

I think my main gripe so far is the UI. Everything's a bit sluggish and many things need 1 click too many. Inventory being the worst offenders. Also some stuff feels sorta hidden, or the game doesn't really give good feedback so you know what is happening and why (like for example when you lose items after a battle). But of course safe to assume there'll be changes along as stuff gets finalized ^^;

Good luck with the project. Seems ambitious, but could be nice.

Yeah, I totally agree with you.
The UI and ergonomy of the game is still very much lacking but it is indeed the last thing I'll be working on, since I have more features to implement beforehand.
It seems indeed very ambitious but I believe I can manage it, by staying very simple in the ideas I've exposed.

I might have a suggestion for you. But of course do whatever you want. It also might very well not be compatible with your current setup or might be to big of a change, but I feel like it could serve as inspiration. I am also not sure how well it would work performance wise.

I feel like the aspect of randomness can give replay value. I had a brief look at your game and all the NPCs governed by some discrete states e.g. where they are and what they are doing. A possible idea to make the characters distinct while retaining randomness might be the following.
Say a configuration of the system consists of the locations and lvls of all characters and what they are doing. On top of this each of your characters might have internal values describing their personality.

Than one might be able set up Markov chain Monte Carlo scheme to transition from one configuration to the next. I assume you have some "energy function" being the sum of how well each character is doing (the quality of your city could be included in this as well, as I saw in your devlog that you wanted the city to be variable as well). This energy function could also be higher or lower based on if characters "like" to be in this situation due their personality. So this would assign an energy value to each configuration. Then a transition rate from one configuration to the next could be described by the Boltzmann factor of your difference in energy between the states, as in the Metroplis-Hasting algorithm. (So a change from your current configuration to a randomly chosen different configuration would always be accepted if the energy decreases but if it increases it might still happen based on the Boltmann factor.)
So this transition would be accepted or rejected based on some random number which is higher or lower than its correspond transtion rate.
(More explantion at the Metroplis-Hasting algorithm wikipedia page)
Of course most states should be unreachable from any current one, so one could optimize here. And many decisions of NPCs are independent so these could be performed in parallel.

So this would minimize the energy function while admitting fluctuations due to randomness. I'd assume their would be a lot of metastability, so one could end up with "stationary" configurations which look quite different from one another based on the chosen seeds for the RNG.

Holy dang. You're way out of my league my dear friend xD. I don't get all that theory so I can't follow up.
I'm staying very simple in the design and in the way the NPCs behave.
What I have in mind is giving NPCs needs with a "Mental State" to manage. Their actions will be dictated by their needs first. If they don't have any needs, they will be moving around the city randomly and interacts with its surrounding.

Now, for a city to operate in accordance to its inhabitants, it will just checks the inhabitants needs.
Imagine 10 NPCs. 7 out of the 10 NPCs need a home. The City will very much likely build a home shelters.

You're making me think as well but the Job Locations will also need some kind of variable so the city knows, wether or not, the Job Location lacks something.
For example, a Bakery doesn't produce Pies for an amount of time because it is lacking the resources.
The city will likely build something to respond to that need.

I'm staying very simple with my designs for many reasons, mostly because I'm an uncultured swine dumbass.
I couldn't understand your explanations completly but I get some of the ideas. I really appreciate the time you've spent writing this.
Thanks!

 

[brad thundercock]

hello, cant save. too many errors. what do?

 

[Silac]

hello, cant save. too many errors. what do?

If you're loading saves from previous builds, then it's only normal. You have to start a whole new save.
If you have trouble or getting errors while saving with the new build, do you telling me what the error is?
Is it a white fulltrack screen or a crash?

 

[brad thundercock]

If you're loading saves from previous builds, then it's only normal. You have to start a whole new save.
If you have trouble or getting errors while saving with the new build, do you telling me what the error is?
Is it a white fulltrack screen or a crash?

oh nah homie i just downloaded the game. its actually quite fun, you know damn well how to write diologue. But i had to restart the game from working save files for like 3 times by now, "An exception has accured" , says maximum recursion depth exceeded while picking an object. dunno wtf that is. anyway real good game, i like the girls, the catgirl with the animation is super cute. i hope in the years to come you make a fun deal out of this game.

 

[Silac]

Dang. Could you send me a screenshot of the last few lines of the error please? It would help me fixing that one for future builds.
But thanks for the comment though! Once I'm done with all the coding part, I'll finally focus on more art and writing content.

 

[brad thundercock]

Dang. Could you send me a screenshot of the last few lines of the error please? It would help me fixing that one for future builds.
But thanks for the comment though! Once I'm done with all the coding part, I'll finally focus on more art and writing content.

shit i already closed the app. gomen...

 

[Silac]

shit i already closed the app. gomen...

It's okay! Thanks for reporting me this bug atleast

 

[Lebeaubrun]

ngl the whole interface is kinda scary, the art is really nice tho wish this was a simpler game. Just want more scenes with the demon girl and a bit of story to give it minimal context.

 

[sluttica95]

Can't save the game. "Maximum recursion depth exceeded."

 

[Silac]

ngl the whole interface is kinda scary, the art is really nice tho wish this was a simpler game. Just want more scenes with the demon girl and a bit of story to give it minimal context.

Yeah, I still have a lot of work on all of these aspects. But until 0.1.0, the updates will be mainly about features and testing gameplay loop.

The 0.1.0 will have a complete character route with new sex scenes etc

 

[Silac]

Can't save the game. "Maximum recursion depth exceeded."

Dang. I will need a screenshot of the fulltrack error so I can look into it.

 

[CloneClown]

I ve gotta start reading post description, because it is another game thats early in dev(looking good btw) that has yet to add scenes and i was thinking i am doing something wrong

 

[GU1]

Well, there's only one sex scene, which is the one in the beginning, with the Demon girl masturbating the player. As an avid NSFW player, this wouldn't be enough for me to fap personally xD.

Until 0.1.0, the build updates will only add core features of My World Isekai.
For the 0.0.8, I will work the map Evolution with the "breeding" system, pregnancy so NPCs can have kids with each other. I also want to add "dynamic" quests/events like Battle tournaments and such.
I want to give activities outside the main narrative for the player to enjoy. Just like a Sandbox game.
There's also the crafting system I want to deepen with customizable weapons.

So no new sex scenes or very few characters development until 0.1.0.

Dang was hoping for some new scenes when I saw it was updated.

Keep up the good work Silac, hope to see some new scenes soon.

 

[ChronicPasswordForgettor]

Cute art, horrid engine choice; please reconsider.

 

[Silac]

Cute art, horrid engine choice; please reconsider.

I agree. If I knew the game would turn out like this, I would have went for Godot or something more adequate.
But I already went too far with it :x

My next project will be on Godot for sure.

 

[ChronicPasswordForgettor]

But I already went too far with it :x

I disagree, it's tedious work, but it's never too far to refactor code for another language. It's much easier to do when it's 0.1 vs something like 0.95, and the code structure is there, just has to be rewritten for another language. I think that'd take off a lot of the sluggishness people complain about in the menus, which is #1 typical renpy shit.

Alternatively, you could try bashing your head against the wall and try to optimize, but it's like polishing a turd. Renpy is just not good for anything but actual, bona fide VNs. If you want to go down that path, I don't know, try threading stuff maybe and be smart about it.

 

[Silac]

I disagree, it's tedious work, but it's never too far to refactor code for another language. It's much easier to do when it's 0.1 vs something like 0.95, and the code structure is there, just has to be rewritten for another language. I think that'd take off a lot of the sluggishness people complain about in the menus, which is #1 typical renpy shit.

Alternatively, you could try bashing your head against the wall and try to optimize, but it's like polishing a turd. Renpy is just not good for anything but actual, bona fide VNs. If you want to go down that path, I don't know, try threading stuff maybe and be smart about it.

You're very right but refactoring still requires lots of work. I agree that it would be way better in long term, as I don't need to worry about optimizing etc, but I don't feel motivated to do so, as I also have to learn Godot. I know I got too ambitious for this project but I'll finish it as fast as possible, so I can work on my next project, which would be what My World Isekai is but bigger.

 

[maedanuit]

Outstanding game concept! I can't wait for the release!

Are there any similar games I can try in the meantime? I'm looking for something like this game, with a dating sim style + sandbox and RPG elements...

 

[Silac]

Outstanding game concept! I can't wait for the release!

Are there any similar games I can try in the meantime? I'm looking for something like this game, with a dating sim style + sandbox and RPG elements...

I don't know any so If there are, please, also tell me!