Re: Let's make an audio game engine!
@56
For audiogames, I think the lack of complexity in environments is actually a lack of good mapping tools. The current state of the art is a tile editor that you arrow around in, which is fine, except that anything bigger than 20x20 or so becomes unwieldy quickly. I think that you can do better, and one of my goals is to eventually write a semantic map language that lets you describe things at a higher level: "this is a hallway. It connects to a room to the north and south" etc. as a DSL of some sort. You could even bring in something like Cassowary arguably, but I have yet to think through what the UI around using constraints would look like with respect to making something sensible that you could script. But it's still not a terrible idea, if you can automate away the difficulties of having to count tiles and such.
You can totally put pillars and stuff in the room with the player. Swamp has realistic shopping centers with isles and stuff (but tellingly the person who wrote Swamp is sighted, and did the map editing with visual tools). Shadow Rine has lots of smaller features in the middle of rooms, though being a top-down 2D game the screens themselves are square. BK3 is a sidescroller that is as complicated as sighted sidescrollers with respect to level layout even to the point of elevators near the end of the game, and I'm saying that as someone who has played lots of sighted sidescrollers before losing too much of my vision.
Nothing stops you using the BK3/Shadow Rine style camera where you hold G and use the arrow keys to "feel" the map (which works much better in practice than it sounds like it ever should) for a first person 2D game like Shades of Doom, where you present a top-down map to the player. Just no one's done it for some reason. Nothing stops you having fountains or something, if you attach sounds to them and have the necessary bits to make it obvious when the player is close.
But as I said above, the state of the art for map editors is you arrow around in something like an Excel spreadsheet seeing only one tile at a time. And the state of the art for audio is stereo panning without even basic filters because all the good offerings are inaccessible to one degree or another (usually because they require that you use their tooling as well). If you had a good way to edit maps and good audio, it wouldn't be hard to just tag everything with text descriptions. But since tagging with text descriptions is usually that you have to explicitly define boxes by hand on top of a grid of tiles, and the aforementioned lack of good audio options, and you get the current thing where everything is square and there's no real interesting features. My main point is: if your map editing doesn't start as a semantic exercise or with sighted-level vision, it won't be semantic enough for presenting complexity.
It's just a lack of programming skill, time, money, etc etc etc. that means that the audiogaming circles don't have the resources to pull all of this together. Plus for the old stuff like Shades of Doom, etc. that was kind of the beginning of audiogaming history and no one had innovated much yet. I'm not overly sure what happened though, in general, because you'd think that we'd have 20 or 30 Shades of Doom clones given that it's been something like 15 years since Shades of Doom and audiogames in general started, but we don't. That has always been surprising to me even accounting for there being no meaningful amount of money in audiogames. We shouldn't be going 3-4 years between something as complicated or more complicated than SoD.
But for 3D none of that is my objection. I agree with your objections to it, and I honestly think most of them are unsolvable. We are even on the verge of having AI narration/description in 5 or so years, at least in my opinion. But even with something like that, as I've said many times in other threads by this point, your sighted friends will have to hold themselves back for you. And what's the point if you can't have fun together, when it's a game and having fun is literally the definition? I can kind of see the point of Second Life given that people there do other things besides gaming, but I have to squint really hard and I will be honest that I have never tried it primarily because of this objection--I could probably learn the tools, but for what, I can just atend online conferences or something instead without bothering if that's what I want to do with my time.
But even supposing you solved that my other objection is that communicating the orientation of objects and the player when the player is free to rotate on all 3 axis and move in all 3 dimensions is really kind of unsolvable. Even in real life when you have the best HRTF imaginable and the inner ear helping you that's difficult. And that's even before you get to "You are on a bridge and the interesting stuff is below you" sorts of problems. You could just put in pathfinding or something but at that point the player isn't even playing the game anymore, the game is playing itself, and now not only have you removed the word fun from the definition of game, you've also removed the word player all together, and just...no. No, no, no.
And then when you realize that halfway through your project in which you decided to use Bullet or Ode, probably around the time you realized that you can't use Bullet because debugging it without graphics is basically impossible, you come to realize that you've just added something like 6 or 7 constraints per object because you needed to turn off a bunch of degrees of freedom so that your limited tooling for communicating 3D environments doesn't break horribly and will be fighting this for the rest of the life of the project.
The only thing I've come up with that might work, which I kind of want to prototype one day, is to use voxels and a semantic map language to describe them, so that you can for example have audio trails that guide players up and down staircases because the engine knows that they're staircases. But you'd never retrofit that onto something sighted or give it graphics appealing to sighted people, and frankly it goes well beyond just tagging objects because you also have to tag floors and such so that you can do things like scan down from the player to figure out what region the floor the player is above is in and etc. And then I stop because I realize that probably less than 10% of the people interested in playing an audiogame would ever be able or willing to put in the time investment to learn to play it and go back to solving the 3D audio problem and thinking about making another Wayfar 1444 but not as a moo or something instead.
@55
You will inevitably learn that I and those of us sharing my opinion are right, but you're apparently going to do it the long way. I would rather have seen you make games, but instead you're going to make something the community doesn't really need because you don't understand what people need, or you're going to fail when you realize that I'm properly estimating the scope of the project.
My main argument isn't "creating a window is hard" anyway, and hasn't ever been. If that's all you've taken away from what I'm saying it's not worth continuing because either the language barrier is in the way or you're intentionally misunderstanding what I'm saying because you don't want to hear it. Creating a window is indeed easy. But a window doesn't make an engine. If you think that creating a window is most of your project, then you either don't understand the scope or you're planning something so trivial that even calling it an engine is a stretch.
If it was so easy that you just had to use SDL and you could have a cross-platform engine in a week because it's 2020 and 2020 somehow makes this easier and anyone who thinks otherwise has outdated programming knowledge, there's at least 10 of us who would have already done it, and probably quite a lot more given that the community has wanted something like this since at least 2012. Another important question you should be asking yourself when you think about starting a project is "why doesn't this already exist?" I can name at least 5 people on here who would have happily made this if it could be done in a week or something, and I know several more who aren't on audiogames.net as well. And the best part is that at least 3 of those people aren't employed for one reason or another and have an infinite amount of free time to do it in and it still doesn't exist. Nothing makes you as special as you seem to think you are. Sighted programmers don't magically have an easier time at this either.
But I'm telling you things you don't want to hear, and you're too new to realize maybe they're true, so of course you shoot the messenger. Fine. Your life, your time to waste.
-- Audiogames-reflector mailing list Audiogames-reflector@sabahattin-gucukoglu.com https://sabahattin-gucukoglu.com/cgi-bin/mailman/listinfo/audiogames-reflector
