Here I am again, with another blog. Don't worry, the section about web development will be small this time, and there will be pictures of my games further down…
Short section on web developmentYes, I did read your comments in the last blog (Thanks Flashback, AG, and sirXemic. I took your advice seriously).So long story short, the site is starting to look more respectable, I'm feeling stressed, and… nothing else.OK. Let's move on, this is getting stale.Long section on my RPG4D entrySo, as mentioned in my last blog, I'm taking a new direction. Or a hundred of them. I hate being so undecided, but I'm admittedly flailing around for solid ground in a genre I've never made a game in (Only a few minor WIP's).And I'm sick of making platformers, so… Yeah.Anyway, I've been experimenting with various ideas, and decided to try create a Roguelike of sorts. Here are the results 20 minutes in:
I.. err… borrowed the Dwarf Fortress font. Interesting fact: This little 'game' builds upon my code from Arbiter (Which I turned into a framework, it was so awesome). The reason I took 20 minutes to make this was because I was trying to create a data structure that could handle multiple 'objects' lying on the 'ground', with a large world, and without lag. Anyway, I was working on that, when I came upon and old game of mine which some of you will remember (And many more won't) that I was working on in 2010:
Damn, I'd forgotten about this… let me just take a look at the project source code quic…. AAAGH MY EYES…*closes Code::Blocks*Ah, now I remember why I cancelled that. My 'engine' (Scourge) was just plain bad. But it was an interesting idea I had (First person roguelike).So I again started yet another game in my RPG4D folder, this time for a little 3D game, which is going to use Minecrafty style blocks for it's terrain and dungeons, but is going to (hopefully) be a roguelike, in first person, in real-time. So another 20 minutes down the line and I had these:
Some of my own textures
Rendering each and every block texture in Minecraft, ad nauseum
Fear not, the second shot isn't actually rendering all 4096 blocks. It's using a basic "Am I surrounded?" test and a basic occlusion test to speed things up. And it's fast.I'm probably going to use Octrees to store the data, because that's logical (Considering I'm dealing with absolutely perfect geometry for Octrees, it should be very fast too).Visibility determination is going to prove to be a pain in the rear, as usual, but I can employ one of the many algorithms I've picked up for this cause. I'm considering an alternative method that I thought up earlier as well: Render all blocks on the top or bottom of the chunk (depending on player orientation in relativity to the chunk). Then, trace through all the empty blocks, and render any immediately adjacent blocks.A test from the empty block's position to the player could be determined as well, if necessary. (This is using the fact that a player can only see a block if it is adjacent to an empty block. Any block not adjacent to at least one empty block is covered up completely).Anyway, this won't be Minecraft, or anything like it. Any 'building mode' will probably only be used for level editing. So swords 'n sorcery, dungeons, towns, NPC's and swag. well, let's see how far I get with this. Start making bets now on what game I'll release, if any, for the comp…Oh, by the way… this game is also using the exact same basecode as Arbiter. For once I've made a really useful (For me, anyway) piece of code. And it mixes 2D and 3D perfectly.Well, back to my sketchpad of many ideas and little progress…
I cast magic missile into the darkness!
I don't understand any of this stuff that you're talking about, except occlusion/hiding stuff, but I like that second screenshot.I think you should just render the first roguelike idea in the 3d style of the second roguelike idea, ie very simple grid-based movement, no animation. You're probably already tempted to make an "Elder Scrolls" style game, but you should really consider just how much work you're taking on with that, as it's fairly easy to underestimate. I think you should think of a compromise which allows you to make a 3d game with the simplicity of a text-based game. Your two ideas are on complete opposite ends of the spectrum and you want something in-between.
sounds cool. I'm guessing randomly generated world. this is going to be a lot of work so you better getting moving quick. even if you bust your ass on this, it's gonna turn out to be limited and small, that kinda project is huge - still it's possible to get a good short game out of it. i hope you stick with it though and good luck.
(now to drag my ass back to GM and work on my project…)