Hi all! I'm the developer of Harling, a fantasy MMORPG I've been building on my own. It runs natively on Linux: the launcher comes as a .deb for Ubuntu and Debian and as an AppImage for every other distro, and once installed it keeps the game updated by itself.
The world is Orand, a continent of 58 regions (these tests will focus on just one of them) where four ancient Labyrinths are starting to give way. There are six classes (Knight, Mage, Archer, Barbarian, Monk, Assassin), group dungeons and voice chat with the players around you.
Harling is in an early stage of development and still in testing, so not everything works well yet: expect bugs, balance that needs adjusting, graphics to improve and more. I'm looking for testers who enjoy playing and telling me what works and what doesn't. Spots are limited and invites go out in small waves. The server is in Europe, so the ping is best from here.
Sign up here: https://playharling.com/#partecipa
I'll be in the comments for any question, Linux ones especially. Thanks for reading!
Edit (Oct 12): in-game screenshots, only cropped to hide the HUD.
Update (Oct 16): a first gameplay video is up: https://youtu.be/DPfqgG2HInQ
You have a good concept, but not really a game.
If you're willing to transfer most (nearly all) of the workload to be non-LLM, message me and I'd be willing to help.
But know it is not industry standard to use LLMs for games. They do not write optimized code, which is the last thing you want for a game, let alone an MMO.
I can also help build you a basic lightweight website you could self host if you'd like.
Thanks, I appreciate the offer and the honest feedback. I’d politely disagree on one point: AI is being used across pretty much every field now, and I think it will soon be hard to compete without it. The real question isn’t whether it’s used, but how. I’m pretty new to game development, but not at all new to programming, and nothing generated goes in without being reviewed, measured and, if it's needed, reworked. Without a solid grasp of architecture and software design best practices, and a good understanding of the mechanics, AI can’t do much on its own.
The website does what it needs to for now. Where I’d really welcome human hands is exactly where I’d most like to move away from AI: art direction and polish, world design, mob design, instances, VFX, and sounds and SFX in general. Most of the sounds and some model are actually CC0 packs, not AI, but they’re not at the quality I’m aiming for. If any of that is your thing, I’d be glad to hear from you.
I can work with that, mainly world design and SFX. And it's good that you have experience in programming. Just know game development is a different beast, unlike standard software development. I agree it can be useful in some minor things, but from personal experience of just using it to understand the technology, it really mucks things up when it comes to interactions. It's good for repetitive code generation (especially collision) but not for tasks such as enemy AI. As long as you know how to code for game systems though, it can still speed up the process.
You also have to remember that when it comes to making a game, you're not competing with production but you're making a form of interactive art (or, of only for the money, a gambling machine, but that's not what this game looks to be). LLMs won't make much of a difference in the indie field, because you're not a large studio with shareholders that needs to focus on speed and cutting costs. You have time and flexibility, and no shareholders to answer to.