pocchi
@pocchistudio
pixel artist / composer / game dev I'm working on #FalsePrincessIruruu, a Sekiro-like pixel-art game. Wishlist: the other place
godot's scene thumbnail actually showing something useful instead of a random grey pixel took me by surprise lmao
i just LOVE spending hours fixing a release pipeline that I run once a year and can already do locally in 5 minutes
hi-res pixel art study is that the correct term for this? whatever its called, now that I'm trying it it's super fun. I get to draw larger characters without getting hung up on minor details like you do with normal digital art.
I was genuinely considering doing something like this if I hadn't known lmao
just found out today that aseprite has a thing called reference layers where you can superimpose images without it being pixelated this works so much better than just resizing hand-drawn sketches in aseprite where you lose a lot of small details
wowie 5k wishlists... another excuse to waste time drawing for fun instead of finishing the game XD
ok so first i got the dumb idea to have my metroivania locks-and-keys™ be firecracker bombs because i like explosions and then i got the funny idea to make the explosion actually have a hitbox
I'm re-introducing platforming as an exploration mechanic. I've gone back and forth in the past whether or not to do platforming, but after implementing wall-jump and ledge climb, I think it's worth another shot
I guess there IS one gameplay purpose in my case, it's a hint that you should perform a guard counter instead of a deflect, to prevent knockback and keep up the aggression.
doesn't serve much gameplay purpose, but I think these extra flairs are important for games where all you do is deflect. both Sekiro AND Nightreign have unique animation when deflecting strong attacks, and I think it adds a lot to the gamefeel
bandits from the Yunnan region this time. much easier to design and work with. I guess there's a reason bandits are a staple for any historical stuff. as usually i fudged the proportions and everyone looked a bit small relative to iruruu, but oh well ill fix that later
I basically had a script that uses collision shapes to map out shards of the main sprite, and when broken it'll use them to create multiple piece nodes via an AtlasTexture of the original sprite. I think having a pixel viewport and applying some light physics helps sell the effect a lot.
wasnt gonna implement breakable props as I thought it'd be too much effort to draw individual shards for every single props, but turned out generating them from an atlas of the base sprite programmatically doesn't look half bad
here's the real reason I added guard counters: element catching (I need a better name) attacks with elemental/status damage that would otherwise chip and proc can be 'caught' with a guard counter and imbued into your sword, which protects you from further status damage.
guard counter! unlike a deflect, there's a slight delay before the active frame, so it's a bit harder to pull off.
been spending some time creating drafts of the next area to see what i like ...I say "draft" but they'll probably remain in this state for the next year and a half xD
after lots of playtesting I realize why Sekiro and AC6 put posture in the middle, as high-pressure combat really doesn't let you peel your eyes from the fight for even a second. it's even more important now that I've added the disarm mechanic. BUT a centered UI element looks weird in a 2D game
im a little late but, HAPPY NEW YEAR!!!! xD since it's the year of the horse I figure I'd draw young Iruruu when she still had her horse and bow back in Xinjiang
That said, this approach is only useful for games where combat has many branching paths. If there's no parries or counters, that occur regularly, it's likely overkill. However, in a deflection-focused game, a single combat event can have like 10+ results, so it helps to have a pure data flow.
The main advantage of this approach is (1) everything related to how combat is done is just a series of reducers and mappers, so debugging is much easier. There's no state I need to keep track of. (2) unit testing is dead simple. I can mock any struct as input and just check the output.
Haha I actually love discussing technical stuff so I'm more than happy to go into details! I'm using Godot 4 with C#. (I actually had this flowchart a long time ago to help myself visualize the whole flow)