NEOGEAR - Cyberpunk Vertical Bullet Hell

[brettc]

I'm working on a cyberpunk vertical bullet hell shmup called NEOGEAR. It's inspired by Cho Ren Sha 68K, DDP, and the works of Tsutomu Nihei.

CURRENT STATUS: Caravan prototype released; full development beginning soon!

I'll post updates in this thread and keep this original post updated with the latest links and status of development!

On June 10, 2026 I released NEOGEAR -Summer Caravan '26-, a 2-minute time-attack mode for web, Linux, Windows, and macOS. It's a prototype for what I hope will become a full, multi-stage game.



Play the caravan mode: https://brettchalupa.itch.io/neogear-summer-caravan-26

The caravan release features scoring centered on collecting chips dropped by defeating enemies and cancelling bullets.

The key art and character designs are by Chromatic Arcana, and the game design, coding, and sound effects are by me. Music tracks from doranarasi, under their permissive license.

[brettc]

Here's a GIF of the gameplay from the caravan demo:



And here's a video of me playing through the game: https://www.youtube.com/watch?v=nS74zIcQaaE

I learned a lot making the caravan mode. How to improve visibility, balancing and tuning spawning schedules, and working with sprites in shmups. While I have made smaller games before, this is the most complex (and in my opinion, fun). I know I've got a lot more to learn, like how to program dynamic bosses with multiple parts, varied bullet patterns, and scoring/rank.

I made the game using Usagi Engine, which is a free and open source game dev tool I made. You code games in Lua and your game live updates when your code and assets does. This made for a really rapid development experience for the caravan mode. But as I prepare to start working on the full game, I will likely switch to coding it in Rust so that I can have better performance. I also prefer statically-typed programming languages and a compiler. There comes a point with scripting languages like Lua where it gets to be untenable and a hassle to work in for me.

I made my own schedule editor for the caravan demo. It's very utilitarian. Basically at any frame I can place which enemies spawn in a given lane. It's a hassle to work with, so I think I will make something slightly more robust for the ~5 stages I am planning for the full game.

[Kyper]

Neat game, managed to break 600k:
https://www.youtube.com/watch?v=VdKeSDaVYVQ

Unfortunately I don't know much about anything to offer any real feedback, technical/game design or otherwise.

As you pondered on items you want to implement like varied bullet patterns and dynamic bosses, I think you're aware of things like how the enemy spawns are static compared to other caravan games where the next enemy in queue will spawn if you destroy them fast enough. Of course you don't have to match that exact behaviour in other games, but there's an odd bit of dead space at the end if you kill the boss and last enemies fast enough.

I only played the browser version and had no problems. The only real bug I ran into is that if you get a high score but don't make it to the end, it doesn't save. Not sure if intentional, but otherwise it's minor

[brettc]

Incredible score, Kyper, thanks for playing and sharing! Yes, the enemy spawns being static was something I intentionally chose and deviated from the caravan norm. Mostly for learning, as the full game will not be wave-based. I'll make a minor revision to the end so there's some more popcorn at the end, just so it's not dead space like you said.

The no high score if you die is intentional. I was thinking part of that score is surviving until the end. Maybe that's odd for a caravan, not sure actually! But thanks agian for the feedback.

[Ixmucane2]

Something I would like to do in this Caravan stage is grazing enemy bullets, for benefits that could include score, fractional lives or fractional bombs, and extending the chain.

The triangular coins tend to remain uncollected, maybe they could float towards the player (so that most of them are eventually collected) or at least stay on the screen instead of exiting and do something useful (e.g. multiply score or explode when hit)

[brettc]

Something I would like to do in this Caravan stage is grazing enemy bullets, for benefits that could include score, fractional lives or fractional bombs, and extending the chain.

The triangular coins tend to remain uncollected, maybe they could float towards the player (so that most of them are eventually collected) or at least stay on the screen instead of exiting and do something useful (e.g. multiply score or explode when hit)

Thanks for this feedback! I am unfamiliar with grazing, I haven't played any games with it yet. So I'll do some research. I'd like to put that in the full game. Noted!

The chips do float toward player a bit, but it does require you to move toward them as the pull gravity isn't very strong. I wanted to try to balance risk/reward with collecting chips vs handling the oncoming enemies. Love the idea of making them do a bit more.

I think what I will end up doing in the full version of the game is add a caravan stage in addition to the main stages where I can experiment a bit more and try these new ideas.

[Ixmucane2]

Currently chasing coins is so difficult, between their extreme dispersion, the short time before they are lost, and the need to stop and dodge bullets, that it doesn't seem worthwhile.
Since, on the other hand, the current precision dodging action is very good, I suggest reducing the demand for chasing coins rather than making it an easier action in ways that would ruin dodging (less obstacles, faster movement).

"Handling incoming enemies", i.e. shooting, could benefit from a quota of automatically aimed or guided weapons (more freedom of movement) and from tactically satisfying ways to take risks to increase firepower without powerups, like increased damage at close range or chain reactions of large enemy explosions taking out nearby enemies (rewarding the player for herding them into a crowd instead of shooting eagerly).

Another thing that coins could do is acting as obstacles that enemies avoid and cannot pass and/or explode in their face as proximity mines. This doesn't require particularly smart control and aiming, just dispersing in ways that tend towards a uniform distribution over the screen.

[Verticen]

Impressive game and associated game engine. Was impressed with the clean, crisp visuals and intelligent design details such as hostile enemy bullets visually sorted over the player’s autofire bullets.

I thought the game was fun and good. Visuals are primitive but very readable in action. It took me a moment to realize how killing enemies doesn’t actually give you points (directly)
Highscore: 251100

Currently chasing coins is so difficult, between their extreme dispersion, the short time before they are lost, and the need to stop and dodge bullets, that it doesn't seem worthwhile.
Since, on the other hand, the current precision dodging action is very good, I suggest reducing the demand for chasing coins rather than making it an easier action in ways that would ruin dodging (less obstacles, faster movement).

I disagree with this comment; the ‘coins/chips’ are “worthwhile” b/c they are the only game element that awards score?? I think the challenge offered by dodging the attacks alone too easy much of the time without incentive to try to chase point items though the attack

I’ve yet to dive into your Usagi engine extensively yet, but from demonstrations it looks good so far. The game program ran just fine on low-end hardware for me. I’ll def be keeping an eye on the engine as it progresses. Also the ability to export a game project to LÖVE seems a crazy OP feature imo.

While ik the purpose is to have a light, minimal engine, I’m curious how challenging do you think it would be for devs to make a game perform TATE rotation with this engine?

[brettc]

While ik the purpose is to have a light, minimal engine, I’m curious how challenging do you think it would be for devs to make a game perform TATE rotation with this engine?

Thanks for the kind words. In a past version chips were much easier to collect but I made it more difficult to add risk/reward, like you said.

With regards to tate, I am going to add an engine-level feature to rotate the game so that I can have tate for my games that need it. NEOGEAR is square, like Cho Ren Sha, so there's no benefit in tate in terms of screen real estate. But I think it's inevitable that it ends up being in the engine itself.

---

I've gotten lots of thorough feedback on the Summer Caravan '26 demo in various places, which has been so incredibly helpful. To summarize it and my takeaways: enemies are too small, enemies HP is too high, chips too hard to collect, the limited bombs but using them to cancel leads to a weird dynamic that's not very fun, and the scoring is a bit busted and exploiting it to get the highest isn't very fun and doesn't necessarily reward skill play. Also, the time-based spawns is wrong for a caravan mode, it should be largely wave-based where clearing a wave spawns the next wave.

All of this makes sense to me. I'm going to work on revising this for the caravan mode that will be in the full game. I'll also likely change some of the core systems too. This was my first attempt at scoring and caravan, so I've got a lot to learn. I've been studying Soldier Blade and Raging Blasters more lately to really figure out what I want NEOGEAR to be.

As part of taking a break, I wrote a free tutorial on how to code a caravan shmup that covers the basics: https://book.usagiengine.com/02-shoot-em-up.html Might be of interest to some people.

My plan with NEOGEAR is to redo the caravan mode and the core systems and get that polished and release that as a demo. Then work on the stages. Currently planning for 5 stages. The sprite work will be the most challenging part I think. But I should grow a lot by taking the project on. While the current caravan leans more danmaku than PC Engine era, I am actually going to shift the game to be more like the Hudson classics and Raging Blasters, as I enjoy that style more and want to explore it further. But there will be some more intense bullet patterns from bosses and mid-bosses.

Hopefully more to come in the next few weeks!

[Ixmucane2]

While the current caravan leans more danmaku than PC Engine era, I am actually going to shift the game to be more like the Hudson classics and Raging Blasters, as I enjoy that style more and want to explore it further. But there will be some more intense bullet patterns from bosses and mid-bosses.

Danmaku moments can come and go very quickly, and specialized enemies aren't required; multiple minor enemies (below the level of "mid-bosses") can provide both ample space for dodging when alone and bullet mazes when in a group.
For example, medium enemies that shoot in oblique directions can appear in a line formation, fire a tight curtain fire until they start dying, and when the formation dissolves move fast to aim at the player.
Or the opposite: relatively hard to kill enemies can immediately discharge aimed and dangerous shots on appearance (common in, for instance, Raiden games) and then progress to dense bullet patterns if undisturbed (maybe with reinforcements).