0909, 0865, 0776

Effective collaboration takes time so IMO #solodev is the default for those who have a part-time #gamedev hobby and want to finish something. Buying packaged products (e.g. art) is a more efficient solution. Granted, the end result is not as custom. AI will also enable more solodev.

Gain the satisfaction from the building process as well as the finished product. I'm finding that for a self-published game that "build it and they will come" just doesn't seem to apply. I am now enjoying the process of building a following, which I know will take time.

Figure out how an AI assistant can help. Especially if you are more into seeing your design implemented quickly and not necessarily into digging into all the technical details it's a great motivator because you see quicker progress.

Agree. It depends. And for a casual game, I think even 15 minutes is too long.

And I get to reply to your posts on both Bluesky and Mastodon (manually)! The fragmentation is a challenge. I've vibe coded a tool using Ollama to help find interesting posts to streamline my time use with social media. Bluesky and Mastodon have the most available APIs whereas other services do not.

Tooltips are a great benefit for desktop UI. Wish there was something as usable on mobile. Maybe mobile screen will detect hover vs. touch some day....

Do the LLM options have decent knowledge of Vulkan? I've found that even if I want to learn something directly having the LLM do the initial exploration really saves time. With some web APIs (e.g. Bluesky), I haven't even had to touch code with a vibe coding solution like Replit.

AI is great for building quick tools that only devs will use.

Having a game loop working early the you can enhance is motivating. Try not to leave your game in a broken state. "Art is never finished, only abandoned" - Leonardo da Vinci (it's evidentially debated if he ever wrote or said this, but it is a good quote anyway)

I went that way. Had a multi-player game working but the session management and scaling were a dev time sink. Decided to go with a simulated MP with AI opponents instead. NPC-AI managed on the server for at least some IP protection in a browser game. #webdev #indiedev #gamedev

Initially for fun in college. And now as a hobby. I'm not sure that makes me "in the games industry" though. To be honest, and not to be negative, game development seems like a very challenging way to make a living. Finding a large capital source would seem to be the way to go if that's your goal.

I like itch.io for browser-based games. Haven't thought of tackling steam but if I had an installable game would head that way.

Vanish, though there isn't that much difference to my eye.

With some practice, I've found that the productivity boost with LLMs makes game projects more doable.

I've been finding that using an AI assistant makes these projects go so much faster that I don't get worn out with the details.

My experience so far has been "build it and they will come" probably doesn't happen, so I've been working on an app to help me keep up with Bluesky using the API to pull posts and an LLM to analyze the posts. Trying to manage my social media time, but it's where the audience resides.

I've been enjoying creating smaller games that are steps along the way to a bigger game maybe some day. Vibe coding on Replit has made this very practical. This single-player game taught me all about desktop/mobile and server component for IP protection. solar-dominion.pastime.games