Did not update the devlog these past few days but I did work on the project. I've done the greyboxing for a cave level using the assets that I made last week and currently I'm about halfway through completion of this first cave level.

Here is how it looks now:

At face value, that doesn't say much. But if I were to add a reference near it, then it starts to say something. Here's how it looks with a character reference:

OK, now we're getting somewhere with this. You get the scale of everything in relation to the character, at least.

I also did the logic for one of these sections. It applies the same logic as the collectibles: there are 4 spawn points and at runtime a switch will spawn at one of the spawn points, randomly.

Initially, I wanted to apply some controlled procgen by way of room spawning: at runtime, in certain places, a room will spawn that is different every time and will lead to different paths. But the way I've laid out this level does not lend to this feature, because this level was thought up as an experience from start to it's current state. So I'll take note of this and do the controlled procgen in the next cave/mine level.

After that, I thought of blocking some exits as a controlled procgen thing, so the experience would be "OK, this playthroguh you're going this way, the next playthrough you're going the other way". And I thought "Nah, this is not adding to the experience necessarily". And then it struck me. How about having a switch of some sort (for now*) and the player would have to find the switch. But each and every time, it will be in a different place (out of 4).

[i]*I say "for now" because the plan is to have a bundle of dynamite that is blocking the way and the player needs to find the detonator. The placeholder switch works as the detonator, because it does exactly what it's supposed to do: go from OFF state to ON state and possibly add a delay to the explosion so as to not detonate instantly.[/i]

So I stuck with this, but as I'm writing this devlog, it struck me that this is something very close to the switch-hunting that was in 90's FPS games. That is not really a fun feature, so I have to be careful and steer away from switch-hunting.

You might also notice two big rooms in one of the screenshots. Those particular rooms will probably not have the same shape when this is completed. Those shapes are for reference only.

Now, you guys might be curious as to how this level will be set dressed. And to that I say: what a coincidence, I'm curious too. :)) But no, really now. I mentioned in a previous post that the set dress will NOT be "model a variant of the room in blender then import it in the engine".

Well, now I'm not so sure that this pipeline won't happen. Setting aside that placing small-ish individual assets around the rooms might get tedious and time consuming (I actually love doing that, it's relaxing in a weird, twisted kinda way), the process might turn out some repetitive results. You would probably start to notice that I was using only 4-5 different meshes, even if I scale them to ungodly sizes and shapes.

So I'll really need to think this one through. I won't exclude a combination of both modeled rooms + individual assets placed around. But this is a problem for future me to properly solve, first on paper, then in the real world and then again in the game world.

So that is all I have for now. See ya in the next post, bye.