So let's talk about that game...
12 months ago
>> Throw me some love! <<
http://www.patreon.com/cobalt
TRICK'N TREATS is a project borne out of sheer madness, which seems suitable considering.
As some of you may recall, my life sort of fell apart last December and it's been a long slow climb out of that hole. Here it is a year later and I'm only just starting to come to grips with my world again.
Just before it all fell apart, I'd had the idea to throw together a renpy game, spend a few months working on it and put it out as a christmas freebie. That ended up not happening. After the crash of December, the project was put on hold just like everything else in my life. And when I finally went back to it months later, I wasn't satisfied with it. It was basically just a slideshow.
So I stuck it in the junk folder and got on with doing what needed doing.
Over the past year, I haven't really drawn much art. There just hasn't been time for it. But I have managed to grab the odd moment or two here to poke and prod at AI image generation, and create a bunch of LoRA models. Which naturally involved trying to teach it to draw like me. Or at least in a close enough representation of my art style.
Anyway, On the first of October this year an idea flashed into my head - what if I took the concept of that old, unfinished game, and used AI to make something out of it? Specifically by adding some actual gameplay to it so it wasn't just a slideshow.
And it's October, so let's make it Halloween-themed. That gives me an angle I can hang the whole design on.
But it was the first of the month. And Halloween was 30 days away. Can you make an entire Renpy game in just a month?
There's a sort of "iron-man" challenge among renpy designers, where they try to do just that. But it's something that a lot of them fail to do, even working with a whole team.
But, I thought maybe.... by kitbashing some of the code out of my older unfinished projects, using Ai to generate the artwork and music, and working more as a producer than an artist... could I corral all these AI elements together into a basic but payable game?
I decided to give it a shot. It was something I could actually do in my (very limited) spare time, or so I rationalized it.
So... gameplay. I used an online AI maze generator to generate me a 10 x 10 maze. I then took the result and changed a few walls around to make it more suitable for a dungeon crawl. (I wanted to make sure the "correct path" covered as much of the floorspace of the map as possible)
In renpy, you can jump to what's known as a label, where a bit of code can be played out, and then you can jump to another label, and so forth. So I created 100 "labels" each to represent a grid square on the map. And I coded up a compass with buttons to navigate through them. Added locations for random "wandering" monster events. And added "levels" which would up the number and intensity of those events as the game progressed.
In the meantime, I had stable diffusion churning out hundreds of images - something I didn't need to be there for it to do. I could set it running before going to bed, and the next day review what it had given me. Picking through the images let me select the good ones and discard the ones with anatomy errors.
But the thing that I had to come to terms with the most to get the project together was cutting back on my design plan. I had planned on, and designed in space for a massive quantity of content which would be randomly selected from. I ended up with only just enough to make the game playable instead. Time was short, and that meant sacrificing my grand plans for "hey, it works."
While this was going on, I poked around on SUNO, making use of those 50 free song generations per day to experiment with different mixes of keywords to generate the specific styles of music I had envisioned. A MUNSTERS-esque opening theme, some spooky filler music, and some "sexy 80's porn music" with a slight "Thriller" feel to it for monster events.
Even cutting corners, reducing my plans, and using AI, and kitbashed code from other projects still took up what little free time I had left to me, but I somehow actually managed to do it. I managed to release an (admittedly buggy) version to my patreon patrons (and other folks) over on my discord just before Trick Or Treat time began locally.
And now, here we are, after midnight. I'm finally releasing a slightly updated (bugs fixed) version here and on Inkbunny.
It's still lacking a few things I had wanted to add, but it was designed with space inside to add more to it should I revisit it again in future.
For now, it is what it is. An entire renpy game. In one month. And I hope you enjoy it.
And again, if you'd like to help see more things like this get made, my patreon and ko-fi are linked in the header of my account page here on FA.
Cheers all.
As some of you may recall, my life sort of fell apart last December and it's been a long slow climb out of that hole. Here it is a year later and I'm only just starting to come to grips with my world again.
Just before it all fell apart, I'd had the idea to throw together a renpy game, spend a few months working on it and put it out as a christmas freebie. That ended up not happening. After the crash of December, the project was put on hold just like everything else in my life. And when I finally went back to it months later, I wasn't satisfied with it. It was basically just a slideshow.
So I stuck it in the junk folder and got on with doing what needed doing.
Over the past year, I haven't really drawn much art. There just hasn't been time for it. But I have managed to grab the odd moment or two here to poke and prod at AI image generation, and create a bunch of LoRA models. Which naturally involved trying to teach it to draw like me. Or at least in a close enough representation of my art style.
Anyway, On the first of October this year an idea flashed into my head - what if I took the concept of that old, unfinished game, and used AI to make something out of it? Specifically by adding some actual gameplay to it so it wasn't just a slideshow.
And it's October, so let's make it Halloween-themed. That gives me an angle I can hang the whole design on.
But it was the first of the month. And Halloween was 30 days away. Can you make an entire Renpy game in just a month?
There's a sort of "iron-man" challenge among renpy designers, where they try to do just that. But it's something that a lot of them fail to do, even working with a whole team.
But, I thought maybe.... by kitbashing some of the code out of my older unfinished projects, using Ai to generate the artwork and music, and working more as a producer than an artist... could I corral all these AI elements together into a basic but payable game?
I decided to give it a shot. It was something I could actually do in my (very limited) spare time, or so I rationalized it.
So... gameplay. I used an online AI maze generator to generate me a 10 x 10 maze. I then took the result and changed a few walls around to make it more suitable for a dungeon crawl. (I wanted to make sure the "correct path" covered as much of the floorspace of the map as possible)
In renpy, you can jump to what's known as a label, where a bit of code can be played out, and then you can jump to another label, and so forth. So I created 100 "labels" each to represent a grid square on the map. And I coded up a compass with buttons to navigate through them. Added locations for random "wandering" monster events. And added "levels" which would up the number and intensity of those events as the game progressed.
In the meantime, I had stable diffusion churning out hundreds of images - something I didn't need to be there for it to do. I could set it running before going to bed, and the next day review what it had given me. Picking through the images let me select the good ones and discard the ones with anatomy errors.
But the thing that I had to come to terms with the most to get the project together was cutting back on my design plan. I had planned on, and designed in space for a massive quantity of content which would be randomly selected from. I ended up with only just enough to make the game playable instead. Time was short, and that meant sacrificing my grand plans for "hey, it works."
While this was going on, I poked around on SUNO, making use of those 50 free song generations per day to experiment with different mixes of keywords to generate the specific styles of music I had envisioned. A MUNSTERS-esque opening theme, some spooky filler music, and some "sexy 80's porn music" with a slight "Thriller" feel to it for monster events.
Even cutting corners, reducing my plans, and using AI, and kitbashed code from other projects still took up what little free time I had left to me, but I somehow actually managed to do it. I managed to release an (admittedly buggy) version to my patreon patrons (and other folks) over on my discord just before Trick Or Treat time began locally.
And now, here we are, after midnight. I'm finally releasing a slightly updated (bugs fixed) version here and on Inkbunny.
It's still lacking a few things I had wanted to add, but it was designed with space inside to add more to it should I revisit it again in future.
For now, it is what it is. An entire renpy game. In one month. And I hope you enjoy it.
And again, if you'd like to help see more things like this get made, my patreon and ko-fi are linked in the header of my account page here on FA.
Cheers all.
Fire15q
~fire15q
Interesting, very interesting.
Karno
~karno
Sweet! I love it when creatives get...productive? Inspired? You MADE something!
Lokim
~lokim
Looking forward to it
FA+
