Flying Fire

Metroidvania solo developer. Working towards building mod tools to help anyone create worlds quickly and easily.

[Flying Fire @FlyingFireCo]

@thsottiaux When given a long running goal, if you steer, after a context reset a half day later it'll still respond to what you just said as if you just sent it. Causes serious regression in long running goals

[Flying Fire @FlyingFireCo]

Here's some gameplay of the 'after getting ice arrows, the heat rises and the environment for the other areas change' which i was talking about. The heat may look a little too much like lava, but compared to the lava itself it does look different enough so i think it's fine. I really don't have that much left to do for this classic mode world, I've been sitting at 95% uninspired to finish the last 5% and send it out there. I have been working on many other future projects and setting myself up pretty good for doing a LOT of fun stuff I've been really inspired to work on, but the spring of 'project delta' inspiration is running dry: it ran for years for free hahaha but i really need to tap deep into it now. so in the US we have a long weekend coming up and I think that is what i'm going to try to focus on for the weekend. In order to get this project over the finish line (so i can guilt free work on the other stuff i'm excited about) I need to really immerse myself back in Project Delta and find that excitement and inspiration again. Honestly I look back at what I managed to accomplish with the game so far and I am so proud since it is NOT easy to do everything this game is trying to do (not saying it delivers as a gameplay experience, but it's not an empty codebase and that's something to be proud of). Building the world of Project Delta has been the foundation for me to believe in myself as a developer, that I can accomplish something as crazy as a procedural metroidvania + mod tools to make it easier for other people to add to the system. And honestly that sounds crazy in itself, but that has also lead me on the path to everything I'm trying to do next which feels so possible now.📷

in full disclosure i've been experimenting a lot with open ai's codex and trying to see what it can do regarding game development, and it's a little rough around the edges and programs itself into bad edgecases and WAY too task oriented for accomplishing a big goal like a full game, even so but i'm so excited for this new era of game development where there's less boundries between ideation and game creation. Sure there's going to be a lot of bad copy+paste games made, but it also will allow new developers to not have to go through the same really tough hurdles I've had to go through when I had to had write code. But on another level the scary thing is that at least currently, using codex to develop is like trying to tame a wild monsterous beast. It makes the worst development decisions as needed to accomplish the exact task at hand as easy as possible. Over the past couple months I've built an infrastructure around this mitigating these downfalls to help me build many games in the future while still adhering to code that I can read and reason about. It hasn't been easy (and who knows, maybe not necessary once the next generation of ai is released) but i have an engine and workflow which i can essentially work through and build a game. the process is still under development, but the possibility is amazing. At this point AI is basically building mod tools as a byproduct of my process of trying to effectively communicate my intent so it can program it.

My big goal right now is to work on between maybe 5-10 games of different genres at the same time, in order to prove my workflow can smoothly and accurately develop games which are in line with my creative vision (not easy). That's looking like maybe a winter timeline at this point (was hoping for summer). Maybe after that releasing the workflow publicly so anyone can build games as easily: honestly i'm not sure at this point how much of 'this being easy' is my experience at this point and how much is due to the workflow. So at the end of this I'm ecstatic about that the fact that any of this is even possible, but I need to put the seal on project delta before i can move to the next chapter. And the way I feel like i need to do that is find the inspiration. So I might not get much programming done this weekend, but maybe I might do a playtest, or listen to the OST (https://t.co/E937APutz2) to remind myself that all of this work that i've done for Project Delta is amazing within itself (whether its seen as a good game, bad game, or just comp
letely ignored by everyone) and deserves its final polish so I can move on to the stuff that currently has my inspiration.

[Flying Fire @FlyingFireCo]

@godotengine Check out @steipete he seems to have built a real solid process around dealing with this for open claw, if you're looking for advice to a solution that is

[Flying Fire @FlyingFireCo]

@GameChap All they needed to do was crunch the bitrate. You can hear too many details like a high quality mic on the new. The old sounds like radio communication

[Flying Fire @FlyingFireCo]

Alright: Officially back into working on Project Delta as of today. Last week did a little preparation this week actively working on the world. Started with a task i was putting off for a while which is the Ice loop region. Did specific "icy" parallax assets for them. The tough thing with this one is that mechanically it's in the Heat rooms and the lighting is heated the icyness doesn't really come across. Given you get 'ice arrows' here I want the theme to be icy, but going off super's structure it should be a heated section of the caverns.

Actually just thought of this while typing this out: it could be neat to actually do a "heat rise" (like a lava rise) segment after grabbing ice arrows. This way it feels like you grabbed the ice arrows and the heat started after that. That technically removes the soft 'heat lock' on the item which makes sequence breaking a little less fun... But i guess you'd still have to escape from the upper heated rooms if you want to sequence break so it's still a challenge. I think I might do that. Only trying to do small chunks each day to get the development inertia back up. From memory the biggest remaining things are:

1. Icy Caverns -> Hot Caverns transition on Grabbing Ice Arrows
2. Finishing Bestiary entries
3. All tutorial videos
4. Final layer
5. G4 4 Boss Fight Sequence
6. Final Boss (technically i haven't integrated this or the g4 sequence to be mod-tools driven and honestly I don't think i will at this point. Instead of migrating all the logic for these sequences over I'll just keep them hardcoded in their specific rooms)
7. Escape

The 4-7 sequence had me having creative paralysis before since I have no idea where to place it on the map. Though I am feeling much more refreshed now which hopefully will help me think through making a good decision there

[Flying Fire @FlyingFireCo]

@thsottiaux I use it on Windows on the right side of my screen and having the side menu appear every time I move my mouse off the window to the left is a terrible ux because it covers the information I want to see

[Flying Fire @FlyingFireCo]

@growing_daniel I think it's because in order to respect copyright the models are probably trained on ai generated derivatives of stories instead of the original work. So it heavily biases towards talking about a story instead of learning how to write in an authors voice even generically

[Flying Fire @FlyingFireCo]

In the past solo devs were told to work on one main project, stay focused and get it done. In this agentic development era, for a creative mind that constantly finds inspiration: it seems best now to be working on varying games and genres in parallel that way when random inspiration hits there is a wide net for it to actually be realized instead of ending up in a dusty old word doc that will never be looked at again.

[Flying Fire @FlyingFireCo]

@Dimillian Goal completed in 1s 0 tokens used

[Flying Fire @FlyingFireCo]

@CyberKomodo considering i have my own drawing app now to make drawing within my palette easier, i guess this is today's reminder you can just do things lol

[Flying Fire @FlyingFireCo]

@BeanJuiceStudio Strong structure with enough variation to explore

[Flying Fire @FlyingFireCo]

@_kelindi Doesn't mean you had to give up, software engineering pays the bills, game development pays the soul

[Flying Fire @FlyingFireCo]

@willdepue It's been so long since I've heard those words

[Flying Fire @FlyingFireCo]

It ate up all my paid credits during that time which haven't been reset, which is unfortunate since I like having them in my back pocket when I run out of weekly limit which I thought I had (and the codex job was still running going on like 5 hours, without any notification it switched from weekly to paid tokens)

[Flying Fire @FlyingFireCo]

@NicolasZu I think if you put enough time into programming the right engine, you can really harness much more on all the rest. So from my experience doing so AI is much further ahead currently that this post presumes

[Flying Fire @FlyingFireCo]

@yacineMTB Sometimes you have to if you don't want to waste your entire weekly quota on a simple change. It's a choice

[Flying Fire @FlyingFireCo]

Look, i think google is doing some really cool stuff... but i think it needs to stop sending me notifications of my dad's dog that recently passed away because i added its facial recognition + name so i could make an album of it while i grieve. 6 months later it's just getting a little sad and like i'm glad i could make the collection, i don't need to be reminded every week about it.

[Flying Fire @FlyingFireCo]

@yacineMTB Of course! Just who in the hell do you think I am!?

[Flying Fire @FlyingFireCo]

@thomasbrushdev Though I've been through that enough (2003-2023) to know very much how helpful ai is in a development workflow today and I do not miss the old days yet lol

[Flying Fire @FlyingFireCo]

@thomasbrushdev I mean anything built before 2023