next »

[kkairos]

Totems of the Yonder

This is a project I've been working on for the Metroidvania Month 20 jam, which you can find here: https://itch.io/jam/metroidvania-month-20. Having a longstanding love of procedurally-generated things, I've wanted to do something like this for a long time, having been inspired by attempts at such things like Weapon Hacker and A Robot Named Fight, and Chasm--a noble if somewhat sprawling attempt at the concept.

This devlog is admittedly a bit in medias res and will be more of a daily, bite-size (and perhaps in places more detailed) version of the log I've got going over here already: https://www.youtube.com/playlist?list=PLvccjxfjJs1StJsQBnYxw-gp32J80Q4mO

Here's an example of a map and a room generated:

I'm hoping to lean into something of a Nifflas-y vibe (Within a Deep Forest, Knytt) without discounting my own desire to also make something slightly more dangerous-feeling for the player. [Edited the previous sentence a bit and adding this: Obviously those games are games you can die in, but I guess the vibe/level of danger I'm shooting for is more Rogue Legacy level.] But this game won't be an attempt at a rogue-lite or a "rogue-vania". For a game jam that isn't expressly focused on games with permadeath, I think including it would be unnecessarily alienating for folks, in addition to the general design problem of its being difficult to get right and make feel fair. But it would also be perhaps at odds with some of the vibes the game is giving. (Sometimes, and maybe it's not just me, I feel like I get vibes back from the project, like one might do when working on a novel. A lot of it's inside me, but sometimes the art just tells you what it wants to be?) The pixels are also mine, but the music (and probably some of the room designs) won't be me; that's a friend from discord named 909crime.

I do have a ways to go considering the jam is slightly over half-over, but I'm hoping even if it's not winning material I can make something interesting and mechanically solid.

Have another couple of gameplay shots for good measure:

[kkairos]

So some progress was made today--more cheating with the procgen guiding the generator along to a coherent map structure. Got it placing "special rooms" (well, one, anyway) and generating the surface area, which looks much more like a standard platformer, with little to no issue (after what feels like several issues upon several getting it to work correctly.)

Today's hard lesson, and one I really should've learned by now: If you think your generator will need 100 rules to work correctly, rest assured it will in fact need 200.

Still going to create some more variety in the surface, and I'm way behind on implementing any combat. And also on the realization that ultimately, *not* going single-screen might have given us more flexibility for this project. Probably too late to change my mind on that one, though, and the progress today getting the surface to generate correctly will almost certainly help with making some other items happen, namely, biome transitions, especially to the water area. With the necessary changes, at least.

Also, and this is relatively minor: I moved the horizontal exits up and made them all a little taller. Map traversal was feeling kind of slow and head-bump-tastic and I think that this will help facilitate smoother room designs that are meant to be crossed more quickly.

If I were doing this again, I'd probably look more into slopes than I did, and quite possibly would be using 16x16 tiles with no half-rooms (or with scrolling.)

[kkairos]

Some new screens!

That screen on top is breaking reality (as is a part on the bottom, though this might be less obvious), so the template or its interpretation needs more work. To be honest, I wish I had the freedom of not doing this particular thing for a game jam (or having frontloaded a lot more pgc work than I did) so that I could lean into that weirdness and let the generator lead, but I think that would probably need to be for another project with some different room generation and linking assumptions.

But I think the water template framing is closer to done. Also spikes can hurt the player now, but that doesn't seem quite worth a screenshot.

[kkairos]

Hey all! I didn't mention it here because I fell off of dev-logging, but I did finish a prototype for the jam. We're working on more for the Super Metroidvania Month version, so in addition to making another actual YouTube devlog, I will try to keep up updating here again.

https://kkairos.itch.io/totems

I recommend desktop for the best experience. Weird and unpredictable things possibly related to my lack of optimization or other weird coding choices seem to happen on web.

I'm not moving this to Playtest status yet, though. Feedback is welcome if anyone wants to play and give some, but it's a got a ways to go.

[nathy after dark]

I love the idea of a procgen Metroidvania because I'm already a big fan of Metroidvania randomizer mods and designing a game with the idea of replayable level generation from the ground up, could be even cooler.