It's part of it but I don't think it's really it. Godot has lots of users who have professional systems programming experience; also it has no tool like blueprints. There are two things that play a big role and boil down to the onboarding experience:
- The comfort and of the tightly integrated GUI editor experience and how extensive the editor is
- The onboarding experience (availability and different forms of learning material, community, etc.)
A lot of people, including programmers by trade, want not to have to worry about building or coordinating any tools.
Add to that that a lot of people just want to reach 60 fps, computers are so fast, games indies want to make are simple...
If a library like raylib ever paired with a really nice editor like flash or game maker, and had extensive learning material like we built for Godot, I'm pretty sure it'd attract a lot of experienced and inexperienced programmers alike. I personally almost regret not knowing about raylib 10 years ago because it'd have been really interesting to do that... Maybe someday!
For maintaining some tree sitter parsers and some parsers of our own, to me tree sitter is like the language server protocol: it's more of a hoop to jump through to provide useful tooling to end users (because it's become a requirement for a number of things).
You could come with a standard API for sharing information with any code editor without tying users to a parser generator; just like you could provide autocomplete and other features without a client and server exchanging loads of JSON on every keystroke.
In practice building the GDScript parser (Godot game engine's game scripting language) for tree sitter is actually a bit harder (much more prone to edge cases/regressions as you edit it, errors are generic and you can't step through your parser with a debugger) than coding a parser yourself. Despite all the effort tree sitter's developers put into building and streamlining it (and looks like building this tech has taken loads of work).
There's definitely that, also taking the time to explore different points of views, doing little thought experiments, and thinking against oneself.
You could have someone focused exclusively on the problems of AI, which wouldn't skip issues, but also not be nuanced (even if it the person's informed and interesting to read, like some of what you might read from Ed Zitron for example).
The discussions about AI between @cmuratori and @DemetriSpanos are always insightful in ways that's really hard to find out there.
I'm always impressed by the clarity of their ideas and arguments, and their ability to explore the discussions' complex ramifications in a way that remains easy to follow along.
I recommend listening to every episode to anyone who wants to think more deeply about everything that's happening (way too fast) with AI. It's a lot more nuanced and well-thought-out than most discussion out there, and distills insights that are otherwise fragmented and scattered across the web.
New episode of Wading Through AI is out! @DemetriSpanos and I discuss the recent trend of open source projects restricting the use of AI and/or going closed altogether: https://t.co/peUfXUJy6G
@valigo@BenjiGameDev Yes, definitely, plus all the time you put into content doesn't help improve your gamedev skills or business understanding that much. It mostly works out different skills.
True for a lot of stuff, eyes from the right audience on something they're very interested in does something, but there's got to be a fit.
There's similarly a fair number of indie gamedevs who go into making resources for gamedevs in hope to get eyes on their game and it's counterproductive: it's very time consuming, hard to do a good job, and those people are just not potential customers.
@SaladeTomate18@rfleury@cmuratori Probably! I understand the thread as being really more about ethics and OP being interested in how people who use generative AI for commercial work in games think about/would justify these things. Probably looking for food for thought.
At least when OP refers to piracy it's because large bodies of data used for training models was actually pirated. That is to say, that data didn't just come from the public internet, open source software, etc. but from downloading torrents. Like books etc. that are not meant to be publicly available.
There was a news about OpenAI on that a while ago, I think I read the same about Meta, and there was also a class action against Anthropic for which the company had to pay a 1.5B$ settlement.
So if you use generative AI for commercial work, there's a (probably unknown) fraction of the training data that was acquired plainly illegally.
If you make something to be accessible to a broad audience and take it from what the users want and value, wouldn't you tend to be nudged there?
At least in the free software I've participated to, users request it all. You need people with a strong vision and to be willing to say no and leave people out (+ make some users furious along the way).
Working as a committee, you'll be looking to reach some kind of consensus. If you base the consensus on listening to your users and what they say their problems are, to avoid colliding opinions too much, I would expect people to naturally tend to that place.
(Not that that's what I personally want or like)
Some people feel entitled to a lot. Which is your favorite? Mine might be the:
"Its been 6 months, and you've still not fixed it?! WTH are you doing??"
The section presenting @ThisIsDarkDax, @uheartbeast, and our comment as student reviews has been removed, and more detailed credits with the CC license links added. @KenneyNL and @KayLousberg have also been credited (sorry Kenney I didn't recognize your assets initially on the page).
That's now settled.
Hey @StayAtHomeDev and @colosoglobal, I messaged you three days ago asking you to remove my name from the student review section of your paid course immediately. You acknowledged it was wrong on Thursday and promised to fix it. It's still there, three days later, during your launch sale.
You used an old comment I left on a public YouTube video and took it out of its context to make it look like I'm personally recommending your "class" before it's even available. I can't speak for @uheartbeast and @ThisIsDarkDax, but regardless, their comments look like they've been taken out of context too. Apart from being misleading to your customers, it's a false attribution.
You understand this should have been considered an urgent website update, right? Because it's making real people appear as though they said things about your product that they didn't. It's not something you let slide over the weekend during a sales campaign before taking it down.
It's your choice to base your $200 video course on a free and publicly available demo, environment, models and characters that we released under CC-By 4.0. But it's pretty sketchy to vaguely list "assets" as a course perk and a "special gift from StayAtHomeDev". Then it's a whole different level altogether to make it sound like we also endorse the product.
Remove the comment and GDQuest's logo immediately.
There's room for everyone in education, Godot, and gamedev to get honest recommendations and keep competition healthy and ethical.
P.S. Thanks for the creative commons attribution in thin grey font at the bottom of a section dedicated to your own portfolio and only after I mentioned the license. Much appreciated.
I feel like it's a culture thing mainly?
Since we started getting rid of third party libraries and writing as much as we could ourselves in typescript, with more co-located procedural code and more specific names, it's become easy to grep and navigate things. But we use a subset of the language, no OOP, we avoid dependency injection, write larger functions than typical in web code, etc.
Especially dependency injection, it seems to be used a lot in the web world but I've found it often overdone and that it made it harder to trace where things came from. I suppose the same would happen with excessive use of function pointers or void pointers with poor naming?
My comment was not on educational material but on a "Godot news" format. That's why it reads "it's hard to keep up with everything that's happening in the ecosystem" ; I don't watch Godot tutorials.
It was also labeled "student review" as you can see in the picture. I'm not a student, this was not a review.
I don't imagine or judge people's intentions. This thing needed to be removed because it was a false attribution. If you let it slide, it sets a precedent for all creators. As soon as it was removed I had already moved on. It does remind me to be alert and careful, but that's it.
About the game, people can make a comfortable living in this ecosystem without resorting to these sorts of practices. You'll see that tutors making video courses can afford to spend a lot of their time making solo indie games, including the author of this very product.
They may not know it but they really don't need to do things like these, and they won't hurt themselves by properly crediting or supporting the FOSS or creative commons resources they use either.