Log In or Create Account
Back to Blog
DEVELOPMENT

3

1,432
Weekly Update - Collie's Psychepelago Dungeon
3 years ago988 words
I spent this week building the Collie's Psychepelago area! Or at least setting up the tileset.

I posted on my Patreon - ∞ HERE IS A SUBTLE LINK TO THE PATREON THAT I HAVE ∞ - during the week with a few screenshots about the process of coming to a decision about this. I'll just show the end result here:



(...Where the minimap's broken, I just noticed. I actually fixed that, but after taking this screenshot apparently! And it takes like ten minutes to open the game for editing so I can't be bothered taking another.)

I've mentioned a few times that the story takes you through several dream sections, in which you explore mindscapes of the playable characters. Ages ago, I intended these to be just single long battles in an arena - meaning I wouldn't have to worry about stuff like making a tileset or area - and that's what I did with that recent intro attempt, but I felt it wasn't what I wanted it to be. Making them full (if short) dungeons just seems more fitting for something that's primarily a game. RPGs are typically just chains of dungeons with excuses to delve into them, after all.

I wasn't sure how this area - and the other characters' Psychepelagos - should look. They're dream worlds, unbound by normal physics, which should be apparent in some way... but while I'm not restricted by physics, I am restricted by development practicality! I can't do anything too outlandish.

I coded the game's tile-based maps many months ago to use layers of these cliffs with triangular corners. I had to manually select each edge tile and carefully assemble them before, so this week I wrote code for my custom editor so it auto-generates edges, allowing me to make areas like this in just a few seconds:



So that should help.

Due to how the edge calculations work though, there's only one 'base terrain' tile type, with others having to be drawn as patches; I think a lot of tile-based games work this way? Or at least similarly. In the island tileset I have - the one seen in all the other previews - grass is the basic terrain, with sand and cobbles as the patches.

I originally intended for this tileset's main terrain to be (abstractly) rocky, with maybe bizarre additional terrain patches - like blood splotches or something - but I eventually settled on what's essentially a palette swap of the island tileset. So it's something familiar (that is, it looks like recognisable terrain), but made dreamy via an unnatural palette. Works well enough, I hope?

Actually, I'm reminded of this bit from MOTHER 3, where the characters are on drugs and the area becomes what's essentially a dreamscape:



The edge calculation is limited though in that the tiles around the edges must be the base terrain - the red/magenta grass in this case - meaning I can't make the houndstooth 'rocky' bits extend to the edges. The end result's maybe a bit overpoweringly glowy reddish... but maybe that's not inappropriate for a dream world? I don't know; I think it could be acceptable anyway!



I also had to spend some time recoding how collision detections worked. Previously I had custom collision code to prevent the player from walking off edges, but it was unreliable and got stuck a lot, which testers pointed out.

I revised it so that the map drawing code generates invisible walls at the edges and used a Unity character controller instead, so now movement's a whole lot smoother. Good!



I feel like I'm settling on a fairly standard structure which I'm hoping will be fun both to play and to make: a series of discrete little dungeons, each of which has a handful of events and a limited list of monster species. Seems obvious, but earlier what I was doing was more fuzzy; some bits were extended battle scenes, a lot of the main areas were all designed to be combined in a big island...

The thought that I can just tackle the game one dungeon at a time - and that there's nothing that isn't a dungeon - does make it feel a whole lot more digestible!

If this is to be a dungeon, it'll need some monsters. I was thinking maybe there could be between 3 and 5 short battles where Collie's accompanied by Savitr, against a couple of species. I could use the Pawnights I already have, and maybe something inspired by the Monsters - that is, the purple things - from MARDEK and Taming Dreams' intros?

The first thing I drew in 2020 while planning Divine Dreams was this:



And I revised it a bit the other day:



(Which I didn't intend to show, otherwise I would have put a bit more effort in!)

Could work well enough as some generic representation of a 'beast', or the monster Collie believes herself and the Blight Wolves to be!

I already meant for Pierce's Ignorat monster to be a reference to this one from MARDEK, but... ehh.





Also, attention in this blog has continued to drop all year, and I should really be doing more to raise awareness about the project. I'll probably make steps towards that once I've got the first couple of dungeons roughly set up (this intro nightmare, then the island (Sprouting Isle) I've already mostly made), then I can make some video previews that should be representative of the final game. Hopefully that shouldn't be far off!

I'll spend the next week fleshing out this Collie's Psychepelago dungeon.

Oh, and last week I mentioned a Lore page I was working on but hadn't finished yet. I still haven't finished it, but hopefully I'll get around to it at some point!

3 COMMENTS

Falcon64~3Y
These Monsters showing up as the first monsters in the initial "dream" section would definitely be a nice callback to MARDEK! Their imagery seems to meaningfully represent the Blight Wolves (or Collie's interpretation of them, at least) as well.

Also, the concept art of the psychepelago on that Monster sheet seems rather cool! If it's not technically impossible, having giant spikes rise from the void between the islands would perhaps fill the area out a bit, and make the atmosphere more oppressive and nightmarish! What you have right now does work well enough, though, I feel.

Any long-term projects will inevitably experience a loss in popularity and engagement as time goes on. As soon as you release something tangible, interest will likely spark up again.
2
astralwolf92~3Y
Pseudo, what programming language are you using for the game? What languages do you know off? And what is your favourite

I've recently picked up python again and I just got over the OOP hurdle
0
ChuckNorris18~3Y
I just received a link to this area's soundtrack. Loving the 5/8 time signature. The mood is ominous, but enchanting. One of your best pieces.
1
Log in to comment!