Metroidvania Month 5
Deadline: September 15th
Current Progress:
Player Mechanics: Rough Alpha...
Deadline: September 15th
Current Progress:
Player Mechanics: Rough Alpha...
Metroidvania Month 5 devlog
- Thread starter Guessmyname
- Start date
- Original
- Created
- Status
- Ongoing
- Watchers
- 7
- Recent readers
- 0
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
Metroidvania Month 5
Deadline: September 15th
Deadline: September 15th
Current Progress:
Player Mechanics: Rough Alpha (Movement)
World Design: Art Pass'd
Enemy Design: Not Started
Older builds:
Basic Movement Alpha 1
World Greybox Alpha 1
World Greybox Alpha 2
So Metroidvania Month 5 just begun, and I'm in it. The goal is, as you might have guessed, making a metroidvania in a single month.
Because I'm a fan of the old Metroid Prime games, suspect the pixelart 2D metroidvania is oversaturated and am generally not a very intelligent person, I've gone 3D. So far all I've got is basic player movement and a dev 'playground' for testing said movement. You can currently run, jump, and slide/climb walls (whilst jumping off of them).
What the theme for this is going to be yet hasn't been nailed down yet until I've got a good set of mechanics to design levels around. Art-style is definitely going to be Wind Waker inspired though, given the time constraints.
The walljump proved a pretty amusing thing to implement; I started out with a system for wall detection and a 'sliding' state intended to slow the player down whilst sliding on walls to make jumping off of them easier. Then I disabled the vertical speed of that and discovered that with the two combined, it would happily glue you to walls and could zoom around >90 degree corners without flinging you off. Thus, the wallclimbing mechanics were born. I really hadn't expected to work as well or as reliably as it did.
The wallslide has other amusing side-effects too; since it slows vertical movement but not horizontal movement, you can actually use it to cross longer gaps. The jump can clear 4 units and a wallslide around ~5-6.
Current issues are:
- Climbing up onto the top of surfaces (you currently have to vertical jump it)
- Climbing will happily let you 'climb' off the surface and drop you; should probably just prevent you from moving further instead in that case.
- Walljump can occasionally fail to start when chaining (either pressing it too early before you enter the wallslide state or something else; I need to work on that)
- Camera is a goddamn pain in the arse in 'tight' environments or when doing walljump challenges, thanks to the collision detection jumping it around
Then I'll need an enemy and a basic health/combat system. Still not entirely decided on theme yet.
Then I'll start building out the worldmap, with the walljump and wallclimb abilities as movement powerups you have to find first.
If I still have time at the end of all that I'll consider some sort of bossfight, maybe.
Wish me luck!
Feedback on the alpha builds I put up would be very, very appreciated.
Last edited:
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
With the basic mechanics down, I'm now doing some experiments with art style. Theme is... prooobably going to be some sort of 'Lost Woods' / overgrown fae castle thing or something. As I said, I'm aiming for something akin to Wind Waker in style:
Painting is... kind of sloppy if I'm honest but I think it works? It's all done with the in-blender texture painting tools. There's actually only four textures involved; the grassy cliffs, the cave tunnel, the transition between the two (an edit of the grassy cliffs) and a texture for the boulders. The Wind Waker style puts a lot of work on the environment textures and colours, over the comparatively low-poly geometry and rampant uv shenanigans (that cave in the third screenshot is entirely done with one single texture).
Unfortunately tomorrow is going to mostly be eaten up by a medical appointment, so I'm unlikely to have much progress to show then.
Painting is... kind of sloppy if I'm honest but I think it works? It's all done with the in-blender texture painting tools. There's actually only four textures involved; the grassy cliffs, the cave tunnel, the transition between the two (an edit of the grassy cliffs) and a texture for the boulders. The Wind Waker style puts a lot of work on the environment textures and colours, over the comparatively low-poly geometry and rampant uv shenanigans (that cave in the third screenshot is entirely done with one single texture).
Unfortunately tomorrow is going to mostly be eaten up by a medical appointment, so I'm unlikely to have much progress to show then.
- Location
- The Middle of the Howling Wilderness
I played around in the movement playground and it was pretty cool! I'm not really a game dev (or a huge gamer), so I'm not sure how valuable my feedback is, but I noticed that the standard jump seems to kill some of your momentum when you land? It felt a little weird moving around since I'm the kind of person who spams the jump button when I'm walking somewhere. Hope that helps!
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
Ah, that. That was actually a deliberate measure to try and make it easier to land on small platforms (you'd keep jumping, landing... and then flying off again because of momentum).
But yeah, I agree it's not really working. I'll remove it. Player controller kind of needs 'if you move here, x state change will happen' logic in general in a number of places.
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
So in today's episode of 'Guessy is regretting making a 3D platformer'...
Experimenting with layout and blocking stuff out.
One unexpected challenge is that I keep making rooms too small. You can only get a feel for how 'big' a space is ingame by... running around it, which means you need to build it first. And if that room is too small (which it always is) and you need to scale it up (which you always do)... every connected room has to also be shuffled around. It's... tedious. Had to pull the camera in too, it kept getting into collisions indoors making it uncomfortable to work with.
It's also become very quickly apparent I need to implement more mechanics. Ladders (read: 'wallclimb constrained to the vertical axis') and Interactables at the very least, for building shortcuts and such.
Also need to test how well the movement code handles moving platforms at some point. That one's always fun to wrangle.
Experimenting with layout and blocking stuff out.
One unexpected challenge is that I keep making rooms too small. You can only get a feel for how 'big' a space is ingame by... running around it, which means you need to build it first. And if that room is too small (which it always is) and you need to scale it up (which you always do)... every connected room has to also be shuffled around. It's... tedious. Had to pull the camera in too, it kept getting into collisions indoors making it uncomfortable to work with.
It's also become very quickly apparent I need to implement more mechanics. Ladders (read: 'wallclimb constrained to the vertical axis') and Interactables at the very least, for building shortcuts and such.
Also need to test how well the movement code handles moving platforms at some point. That one's always fun to wrangle.
Last edited:
- Location
- 委員長の牢獄
Gonna be keeping an eye on this! I had a lot of fun reading your older dev-posts back when you had a thread about them (which I'm sure you'll hate me for reminding you about 
), and while I'm currenting speedrunning my thesis this week, as you do, I might as well distract myself with something fun~
I haven't touched Unity much since back then, although I did make a game for a class project once that me and one group-mate then ruthlessly recycled for a second class project, which worked very well. Not as a game tho, playing it was pretty awful lol
Maybe I'll get back into it once I graduate or something
), and while I'm currenting speedrunning my thesis this week, as you do, I might as well distract myself with something fun~I haven't touched Unity much since back then, although I did make a game for a class project once that me and one group-mate then ruthlessly recycled for a second class project, which worked very well. Not as a game tho, playing it was pretty awful lol
Maybe I'll get back into it once I graduate or something
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
Gonna be keeping an eye on this! I had a lot of fun reading your older dev-posts back when you had a thread about them (which I'm sure you'll hate me for reminding you about
I haven't touched Unity much since back then, although I did make a game for a class project once that me and one group-mate then ruthlessly recycled for a second class project, which worked very well. Not as a game tho, playing it was pretty awful lol
Maybe I'll get back into it once I graduate or something
Ahaha don't remind me orz
I will, as always, forever recommend the noble art of Just Mucking Around in Unity :|b
For today's progress, I've finished up the world blockout, and gotten ladders and 'Interactables' in.
Ladders are quite literally just the wallclimbing state without the horizontal movement, though with some extra stuff to correctly snap you onto the ladder etc.
Interactables are basically the 'Press E to [Do Thing]' system. Currently the only available interactor lets you drop a ladder as a shortcut from the big tree room to the large starting meadow area. The world blockout is currently kinda small but there's room to expand it, and for now it covers enough to be working with. There's the starting meadow to run around in, leading to three 'tutorial rooms' you have to learn how to jump and climb ladders to pass through, then you reach the tree room and can open the meadows shortcut / oops through a hole into the spider rave pit, collect the axe and hack/jump your way back to the surface.
From here I'll be looking to implement all the other important sundry mechanics, like doors / level loading, health, NPC monsters and powerups.
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
Are you properly including "conspicuously visible treasure that is impossible to reach for a new player in the very earliest parts of the game that you have to backtrack to get once you find the correct power-up" places?
Yep
The upper branches of the Tree Room can only be crossed via wallclimbing. Not that you can actually get there yet as they're dead ends.
There's also a switchback in a twisty tight corridor maze that will get you a health pickup if you happen to spot it.
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
Sorry about the lack of an update yesterday; been working on more 'back-end' features. For example: rooms now load in asynchronously:
Whereas before it was quite literally the blender model export all dumped in one scene, now each room is a prefab, loaded in using Unity's fancy new Addressables API. The idea is that you track the room player is in, and asynchronously load in all the surrounding rooms whilst they're playing. If a room is not the current room or neighbouring to it, it can be safely unloaded to free resources. If the player is trying to open a door before the opposite room has loaded, it'll wait first before opening. Anyone who's played any of the Metroid Prime games is likely familiar with how this works.
The problems and delay stemmed from two fronts:
One, I had to convert every room to an individual prefab and define their exit points whilst still trying to figure out how best to do so that would still let me make updates to them in blender. This got a little tricky and involved an asset post-processor script to automate some features.
Two, the Addressables API is new, I haven't used it before and has some truly unhelpful error messages when anything goes wrong or you use it incorrectly at runtime. To make matters worse, Unity also only recently started supporting async c# code (as opposed to Unity's built-in Coroutine system, which is slower and clunkier). That has even more unparseable and unhelpful error messages whenever that does anything wrong at runtime. By their twisted powers combined I spent several hours making angry snarling noises and trying to find where the hell the most up-to-date Addressables documentation was.
Wrapping my head around the loading system also took a little while, not for the system itself but for how to handle the doors. Obviously, a door connects two rooms. But since we're loading rooms in-and-out willy nilly... who owns the doors? This lead to an approach I cribbed from the Half-Edge Mesh representation model; every Room has 'half-exits', which know the address for loading in the next room they're connected to. When that other room spawns it, will have a corresponding half-exit in the same position, and the Door is shared between them. Whilst this sounds simple on its own it raises the interesting question of 'who gets to decide the door's rotation?' since the half-exits also always point inwards (they double as checkpoints for respawning the player if they fall into a pit or something) so depending on which half-exit spawns the door first it could end up facing in the wrong direction.
The only Door I currently have is symmetrical anyway so this wouldn't matter, but it could prove an issue later, so I had to decide a method of defining a door's rotation from half-exits that would be consistent regardless of which half-exit spawned it.
Tl;dr this is all a wee bit complicated.
Currently I've got loading in and the door prefab half-complete. It needs some more work for the door to actually open and close and for the game to detect when you've crossed through one. Correctly assigning all the half-exits and their opposing rooms was also pretty tedious; I'll be looking to automate that if I start expanding the worldmap later.
That was very... mentally exhausting.
Whereas before it was quite literally the blender model export all dumped in one scene, now each room is a prefab, loaded in using Unity's fancy new Addressables API. The idea is that you track the room player is in, and asynchronously load in all the surrounding rooms whilst they're playing. If a room is not the current room or neighbouring to it, it can be safely unloaded to free resources. If the player is trying to open a door before the opposite room has loaded, it'll wait first before opening. Anyone who's played any of the Metroid Prime games is likely familiar with how this works.
The problems and delay stemmed from two fronts:
One, I had to convert every room to an individual prefab and define their exit points whilst still trying to figure out how best to do so that would still let me make updates to them in blender. This got a little tricky and involved an asset post-processor script to automate some features.
Two, the Addressables API is new, I haven't used it before and has some truly unhelpful error messages when anything goes wrong or you use it incorrectly at runtime. To make matters worse, Unity also only recently started supporting async c# code (as opposed to Unity's built-in Coroutine system, which is slower and clunkier). That has even more unparseable and unhelpful error messages whenever that does anything wrong at runtime. By their twisted powers combined I spent several hours making angry snarling noises and trying to find where the hell the most up-to-date Addressables documentation was.
Wrapping my head around the loading system also took a little while, not for the system itself but for how to handle the doors. Obviously, a door connects two rooms. But since we're loading rooms in-and-out willy nilly... who owns the doors? This lead to an approach I cribbed from the Half-Edge Mesh representation model; every Room has 'half-exits', which know the address for loading in the next room they're connected to. When that other room spawns it, will have a corresponding half-exit in the same position, and the Door is shared between them. Whilst this sounds simple on its own it raises the interesting question of 'who gets to decide the door's rotation?' since the half-exits also always point inwards (they double as checkpoints for respawning the player if they fall into a pit or something) so depending on which half-exit spawns the door first it could end up facing in the wrong direction.
The only Door I currently have is symmetrical anyway so this wouldn't matter, but it could prove an issue later, so I had to decide a method of defining a door's rotation from half-exits that would be consistent regardless of which half-exit spawned it.
Tl;dr this is all a wee bit complicated.
Currently I've got loading in and the door prefab half-complete. It needs some more work for the door to actually open and close and for the game to detect when you've crossed through one. Correctly assigning all the half-exits and their opposing rooms was also pretty tedious; I'll be looking to automate that if I start expanding the worldmap later.
That was very... mentally exhausting.
Last edited:
- Location
- 委員長の牢獄
One of my perennials issues with trying to do things in Unity (and games in general) is that I get started on a concept, run into a half-dozen interesting "if I build the game I need to solve this issue" concerns that obviously have implications for every stage of the design (eg. I had thoughts for a grid-based turn-based system with height and wall-scaling which... yeah. I actually also wanted to add moving terrain lol). I then spend a lot of time hashing out the logically necessary issues and specifications, discover more issues as I do this, start trailing off and eventually get distracted.
And yet I don't think you can actually convince me that my process is inherently wrong (except for the "trails off and gets distracted" part), because there are few things as obnoxious and infuriating as having to restructure your entire code architecture because it has to account for some fucking edge-case that goes to the very heart of your design
And yet I don't think you can actually convince me that my process is inherently wrong (except for the "trails off and gets distracted" part), because there are few things as obnoxious and infuriating as having to restructure your entire code architecture because it has to account for some fucking edge-case that goes to the very heart of your design
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
One of my perennials issues with trying to do things in Unity (and games in general) is that I get started on a concept, run into a half-dozen interesting "if I build the game I need to solve this issue" concerns that obviously have implications for every stage of the design (eg. I had thoughts for a grid-based turn-based system with height and wall-scaling which... yeah. I actually also wanted to add moving terrain lol). I then spend a lot of time hashing out the logically necessary issues and specifications, discover more issues as I do this, start trailing off and eventually get distracted.
And yet I don't think you can actually convince me that my process is inherently wrong (except for the "trails off and gets distracted" part), because there are few things as obnoxious and infuriating as having to restructure your entire code architecture because it has to account for some fucking edge-case that goes to the very heart of your design
Oh god I know exactly what you mean. Right down to the 'turn-based tactics with height and wall-climbing which... yeah. With non-standard cell shapes, connectivity and also per-cell light mechanics because WHAT IS OVERSCOPING.
Cue several, multiple attempts at making a turn-based tactics game, getting completely bogged down in the map system and editor and then getting distracted, starting something else and completely forgetting / being completely unhappy with the old map system and starting again from scratch.
Ad nauseum.
The one big advantage this gamejam has is its deadline. Knowing no matter what, it must be ready in a month (well, just under three weeks now) forces you to focus and cut and not go 'but what if...'
I think the thing I'm learning the most from all this is that deadlines are actually important.
Like, you can think of your limit for getting distracted as its own fuzzy internal deadline in its own sort of way... :Sa
Last edited:
- Location
- 委員長の牢獄
Deadlines are super important!
For instance, I'm writing about half of my thesis text in the last week of the 6 month process because now is when the deadline is here!
I still might try and go back to make that turn-based system at some point though. I reeeeeaaaally like the idea.
EDIT: there's actually some research that suggests that your time to accomplish a task fills out to take up the time you have, so if you tighten your deadlines, you get more done... or if you can manage to free yourself of time-based work and just... work. But that's hard.
For instance, I'm writing about half of my thesis text in the last week of the 6 month process because now is when the deadline is here!
I still might try and go back to make that turn-based system at some point though. I reeeeeaaaally like the idea.
EDIT: there's actually some research that suggests that your time to accomplish a task fills out to take up the time you have, so if you tighten your deadlines, you get more done... or if you can manage to free yourself of time-based work and just... work. But that's hard.
Last edited:
grommile
nasty-minded old cynic
- Location
- NERV-Mercia
- Pronouns
- Traditional/Animate
We call it Parkinson's Law.
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
Oof, again sorry for the delay I was too tired to properly finish this up last night:
Rooms now load/unload asynchronously in the background as you run around the level. Discovered a lot ofheadaches features of the Addressables API along the way.
It works by having assets marked as 'addressable', and then you can use AssetReference objects to point to them. These references do a lot of very useful things, like keep track of how many times an asset has been used and if can safely be released from memory, but...
1. Sometimes, when you boot up the Unity Editor, it will have completely forgotten where all your addressables actually are, even though the assets are still there. !!Fun!!
2. AssetReferences are not directly comparable. If you have two asset references pointing to the same asset, the only way you can tell is if you convert them to strings and compare those. Which becomes important when AssetReferences are the only means you have of checking which rooms you want to keep and which you want to unload when the player changes rooms! More !!Fun!!
There was also some shenanigans with orphaned doors, but that's fixed now. The only truly critical task I need to work on before returning to actual gameplay elements / art stuff is persistent world changes. For instance, there's a ladder in the starting room that can lowered as a shortcut to reach another area... but it won't stay lowered once the room unloads in this current iteration.
Once that and a number of other bugs are resolved, I'll have another build to go up.
Rooms now load/unload asynchronously in the background as you run around the level. Discovered a lot of
It works by having assets marked as 'addressable', and then you can use AssetReference objects to point to them. These references do a lot of very useful things, like keep track of how many times an asset has been used and if can safely be released from memory, but...
1. Sometimes, when you boot up the Unity Editor, it will have completely forgotten where all your addressables actually are, even though the assets are still there. !!Fun!!
2. AssetReferences are not directly comparable. If you have two asset references pointing to the same asset, the only way you can tell is if you convert them to strings and compare those. Which becomes important when AssetReferences are the only means you have of checking which rooms you want to keep and which you want to unload when the player changes rooms! More !!Fun!!
There was also some shenanigans with orphaned doors, but that's fixed now. The only truly critical task I need to work on before returning to actual gameplay elements / art stuff is persistent world changes. For instance, there's a ladder in the starting room that can lowered as a shortcut to reach another area... but it won't stay lowered once the room unloads in this current iteration.
Once that and a number of other bugs are resolved, I'll have another build to go up.
Last edited:
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
I'll playtest sometime after noon on Saturday if there's a build up
Haha, about that...
So after taking four days just implementing doors, persistent world state took all of two hours to get working. Some spit, polish and making the checkpoint system respect what the current room is oops we have our new build.
Last edited:
- Location
- 委員長の牢獄
So as I said, I went and tried it out. My experience of the first attempt goes more or less like:
After beating through my security warnings to actually run it, notice I'm standing on what looks like a great featureless white void that I later realize is the big arena-ish place. I start trying to turn the camera and realize it will only switch a little when I move the mouse, and won't actually look in the opposite direction (for some reason?)
Running around, I'm vaguely annoyed at the footstep sounds, especially the overlap
I start jumping a bit and jump against the wall and see how you can do the slide thing.
I finally proceed to start looking for some part of the actual game, running along the wall to see if I can't find a different room.
I do some jumps against the wall.
I jump too high, somehow managing to catapult myself up past the boundaries and intoSPESS the sky, upon which the game stutters, freezes and dies.
Total playtime: 20 seconds? Finished on force-closing the window.
Second time went better - the camera worked but the sensitivity was garbage. Footstep noises were still kinda annoying but alright.
I ran around the level for a while, reaching the big tree-like thing and then falling into a bunch of holes until I reach the place where you can seriously fall into a hole and restart at the door, which I of course proceeded to do 4 times totally on purpose. I made my way back up to the tree, found my way over to the ladder that you can drop down into the big starting area, and prompting fell off the ledge without activating the ladder, futilely attempted to jump back up despite knowing it would be impossible, and closed the game.
Probably took me 3-4 minutes.
I like the irregular geometry. With some different textures and some props it feels like it could look super nice. The controls are somewhat funky, but not terrible. There were a few moments of stuttering, but I can't pinpoint what they might be related to.
After beating through my security warnings to actually run it, notice I'm standing on what looks like a great featureless white void that I later realize is the big arena-ish place. I start trying to turn the camera and realize it will only switch a little when I move the mouse, and won't actually look in the opposite direction (for some reason?)
Running around, I'm vaguely annoyed at the footstep sounds, especially the overlap
I start jumping a bit and jump against the wall and see how you can do the slide thing.
I finally proceed to start looking for some part of the actual game, running along the wall to see if I can't find a different room.
I do some jumps against the wall.
I jump too high, somehow managing to catapult myself up past the boundaries and into
Total playtime: 20 seconds? Finished on force-closing the window.
Second time went better - the camera worked but the sensitivity was garbage. Footstep noises were still kinda annoying but alright.
I ran around the level for a while, reaching the big tree-like thing and then falling into a bunch of holes until I reach the place where you can seriously fall into a hole and restart at the door, which I of course proceeded to do 4 times totally on purpose. I made my way back up to the tree, found my way over to the ladder that you can drop down into the big starting area, and prompting fell off the ledge without activating the ladder, futilely attempted to jump back up despite knowing it would be impossible, and closed the game.
Probably took me 3-4 minutes.
I like the irregular geometry. With some different textures and some props it feels like it could look super nice. The controls are somewhat funky, but not terrible. There were a few moments of stuttering, but I can't pinpoint what they might be related to.
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
Er, was it literally white? There are stand-in textures but they're grey and orange so that shouldn't have been happening. Even if the world geometry was white there should also be Unity's default procedural skybox so... :S
Could you elaborate on this? I also use the mouse to look around so it could be a sensitivity issue (I'm already going to be adding view control options for inverting the camera stick on gamepad, so mouse sensitivity should probably also go on the list)
Are they too loud? They play in sync with the run animation (ie foot makes contact = play sound and make a little puff sfx)
I jump too high, somehow managing to catapult myself up past the boundaries and intoSPESSthe sky, upon which the game stutters, freezes and dies.
:x will look into
I've had that happen once or twice but I thought it was a profiler bug in the Unity Editor. If it's happening in builds though...
Were you trying to walljump at the time? Both times I'd encountered it involved walls so it would help to whittle things down. Reproducing it's a pain though since it happens irregularly...
Second time went better - the camera worked but the sensitivity was garbage. Footstep noises were still kinda annoying but alright.
I ran around the level for a while, reaching the big tree-like thing and then falling into a bunch of holes until I reach the place where you can seriously fall into a hole and restart at the door, which I of course proceeded to do 4 times totally on purpose. I made my way back up to the tree, found my way over to the ladder that you can drop down into the big starting area, and prompting fell off the ledge without activating the ladder, futilely attempted to jump back up despite knowing it would be impossible, and closed the game.
Probably took me 3-4 minutes.
I like the irregular geometry. With some different textures and some props it feels like it could look super nice. The controls are somewhat funky, but not terrible. There were a few moments of stuttering, but I can't pinpoint what they might be related to.
thanks! Could you elaborate on the 'funky' controls? Were they too sensitive, sluggish or...?
- Location
- 委員長の牢獄
Nah the level was all there, it was just since there wasn't all that much stuff there, it took a few moments for me to orient myself. It's not so much an issue as just me reporting on the confusion of the first run of the build.
I think the camera issue stuff is quite possibly a sensitivity issue. It felt extremely low to me - possibly because I have mouse sensitivity set at fairly low settings for default use, but I basically had to push it all the way into the corners of my screen to get even just 90 degrees turns IIRC. Whether or not there were unrelated issues with the first run is hard to say - it crashed fairly early while I was still experimenting with the controls so!
I think loudness of the footsteps ultimately can't be judged right now - since there's no other sound I perceived them as very loud - but what I noticed was that if I was moving in a way that made the feet impact too often or something, it felt like the footsteps played too often, to the point they overlapped weirdly and you got this "taptaptaptap" noise. I think you should grab some kind of soothing area music and put in first though, because it might just be because it was literally the only sound-response I had.
And yes, I was trying to wall-jump, so it's almost certainly something to do with that.
The controls... I think this might tie into the camera sensitivity more than anything else, but I had issues with the "precision" jumps over the big hole. On a similar note, you sometimes get slightly stuck on doors if you don't jump through them, and when I climbed a ladder the second time I had to jump off at the top - not sure if that's just how it is right now or if it was a bug.
I think the camera issue stuff is quite possibly a sensitivity issue. It felt extremely low to me - possibly because I have mouse sensitivity set at fairly low settings for default use, but I basically had to push it all the way into the corners of my screen to get even just 90 degrees turns IIRC. Whether or not there were unrelated issues with the first run is hard to say - it crashed fairly early while I was still experimenting with the controls so!
I think loudness of the footsteps ultimately can't be judged right now - since there's no other sound I perceived them as very loud - but what I noticed was that if I was moving in a way that made the feet impact too often or something, it felt like the footsteps played too often, to the point they overlapped weirdly and you got this "taptaptaptap" noise. I think you should grab some kind of soothing area music and put in first though, because it might just be because it was literally the only sound-response I had.
And yes, I was trying to wall-jump, so it's almost certainly something to do with that.
The controls... I think this might tie into the camera sensitivity more than anything else, but I had issues with the "precision" jumps over the big hole. On a similar note, you sometimes get slightly stuck on doors if you don't jump through them, and when I climbed a ladder the second time I had to jump off at the top - not sure if that's just how it is right now or if it was a bug.
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
Duly noted :Sa... I have a sound pack so adding ambient music shouldn't actually be that difficult.
With doors and ladders, yeah, I know; the only way off the 'top' of a ladder is to jump it currently. Doors I'm already revising with an art pass, I know they're easy to get stuck on as they are currently.
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
Here's an updated build. Camera control options (press Escape/Start) and improved doors. No luck with the jumping bug yet (I haven't been able to reproduce it so if it's fixed is a big ???).
Working on the character art as well currently; going for a sort of Red Riding Hood theme because it matches the woodlands the game is meant to be in.
This initially involved a cloak, to be animated using Unity's built in cloth physics, but it quickly became aggravatingly apparent that the Unity Cloth is a very special princess who only works in specific, sterilised environments under ideal conditions with spherical cows, and most certainly not on a character exported by blender (you filthy plebian). Like... quite literally, even asides from all the exploding geometry and snarls you get just from adding a cloth component... unity imports blender files rotated 90 degrees no matter how you wrangle the fbx exporter, and cloth defines 'down' in local space.
Guess what happens.
So yeah, that was... several hours wasted. Thanks, Unity :\
Though I did discover a few mistakes in my export settings, correcting which allowed me to get FullIK working properly again at the cost of the player model becoming a terrifying exploded squid creature for a while until I got all my imports and prefabs back in agreement with each other. Nothing in the project is actually using FullIK, admittedly, but I'd been having problems with it outside of this that I'm glad I'll be able to fix later.
Working on the character art as well currently; going for a sort of Red Riding Hood theme because it matches the woodlands the game is meant to be in.
This initially involved a cloak, to be animated using Unity's built in cloth physics, but it quickly became aggravatingly apparent that the Unity Cloth is a very special princess who only works in specific, sterilised environments under ideal conditions with spherical cows, and most certainly not on a character exported by blender (you filthy plebian). Like... quite literally, even asides from all the exploding geometry and snarls you get just from adding a cloth component... unity imports blender files rotated 90 degrees no matter how you wrangle the fbx exporter, and cloth defines 'down' in local space.
Guess what happens.
So yeah, that was... several hours wasted. Thanks, Unity :\
Though I did discover a few mistakes in my export settings, correcting which allowed me to get FullIK working properly again at the cost of the player model becoming a terrifying exploded squid creature for a while until I got all my imports and prefabs back in agreement with each other. Nothing in the project is actually using FullIK, admittedly, but I'd been having problems with it outside of this that I'm glad I'll be able to fix later.
Last edited:
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
Sooo. Hit a bit of a rut on the character / npc side of things trying to decide what exactly they should look like. Since I figured having grey/orange boxes for a world design probably wasn't helping I'm taking a crack at pretty-ing up the rooms a bit. Personally I just find it hard to work on one thing in isolation; there's always that nagging voice in the back of your mind wondering how well it will fit in with everything else.
Initially I tried mocking up some various insectoid like critters meant to just be mook wanderers. Like those crawling bugs you often find at the beginning of Metroid Prime or Hollow Knight that do nothing but passively walk around; a simple, early-game 'threat'.
But then I began to wonder; do giant insects even make sense for the woodland setting? Like, prototyping issues aside (spot the bug that came out looking remarkably headcrab-y)... did it fit? I also experimented with how to place and use AI for these wanderers which lead to yet more Unity adventures with the AI navmesh system, which works perfectly fine and dandy (I've used it before) but for some reason the in-editor navmesh visualisation has stopped working so I spent several hours under the impression I had no navmeshes at all. Fun! I also experimented with having it just move forwards and crawl up walls, but that got into issues with rotations; it all worked fine, but having it snap angles when it changes surfaces looked hideous and trying to blend them looked stupid and compared to just, say, making them follow a spline it just didn't seem worth the effort.
But yeah I'm pretty sure I'm gonna scrap the giant beetle insects. Giant spiders are still... sort of on the table in that the caverns beneath the tree were built with them in mind but thinking on how the heck they're meant to animate and navigate is making me reconsider... :Sa. Ideally, I want something that can just run around on a navmesh floor or doesn't need to move at all. No wall-crawling stuff. Not enough time.
Feeling stumped on that front I went for art assets, figuring I'd need to make progress on there anyway and hoping having a better idea of world would help get the ball rolling again. Unfortunately it turns out designing these things in isolation isn't easy either.
Though I'm much more certain I'll find use for those soon enough.
So I fell back further and starting getting textures from the art prototype in on the level prototype, and I gotta say:
I frickin' love the Windwaker artstyle now.
That room? Is all one texture. One UV map; no splats. The UV shenanigans it took to pull it off are filthy and I absolutely would never get away with it if I went with photorealistic textures. It's beautiful, utterly ridiculous and I love every minute of it. Handcrafting the uv map like this is a little time consuming but it's so much fun.
In hindsight, I should really have been using the central tree room as the art bible on this since it has more actual... stuff, including the actual trees for this forest level I need to model to surround the starting meadow area at some point. But, y'know, these rooms are on the way
Ingame they're suffering a bit because I haven't baked any shadows, and the runtime ones are causing some... issues:
We'll get there eventually I'm sure.
Initially I tried mocking up some various insectoid like critters meant to just be mook wanderers. Like those crawling bugs you often find at the beginning of Metroid Prime or Hollow Knight that do nothing but passively walk around; a simple, early-game 'threat'.
But then I began to wonder; do giant insects even make sense for the woodland setting? Like, prototyping issues aside (spot the bug that came out looking remarkably headcrab-y)... did it fit? I also experimented with how to place and use AI for these wanderers which lead to yet more Unity adventures with the AI navmesh system, which works perfectly fine and dandy (I've used it before) but for some reason the in-editor navmesh visualisation has stopped working so I spent several hours under the impression I had no navmeshes at all. Fun! I also experimented with having it just move forwards and crawl up walls, but that got into issues with rotations; it all worked fine, but having it snap angles when it changes surfaces looked hideous and trying to blend them looked stupid and compared to just, say, making them follow a spline it just didn't seem worth the effort.
But yeah I'm pretty sure I'm gonna scrap the giant beetle insects. Giant spiders are still... sort of on the table in that the caverns beneath the tree were built with them in mind but thinking on how the heck they're meant to animate and navigate is making me reconsider... :Sa. Ideally, I want something that can just run around on a navmesh floor or doesn't need to move at all. No wall-crawling stuff. Not enough time.
Feeling stumped on that front I went for art assets, figuring I'd need to make progress on there anyway and hoping having a better idea of world would help get the ball rolling again. Unfortunately it turns out designing these things in isolation isn't easy either.
Though I'm much more certain I'll find use for those soon enough.
So I fell back further and starting getting textures from the art prototype in on the level prototype, and I gotta say:
I frickin' love the Windwaker artstyle now.
That room? Is all one texture. One UV map; no splats. The UV shenanigans it took to pull it off are filthy and I absolutely would never get away with it if I went with photorealistic textures. It's beautiful, utterly ridiculous and I love every minute of it. Handcrafting the uv map like this is a little time consuming but it's so much fun.
In hindsight, I should really have been using the central tree room as the art bible on this since it has more actual... stuff, including the actual trees for this forest level I need to model to surround the starting meadow area at some point. But, y'know, these rooms are on the way
Ingame they're suffering a bit because I haven't baked any shadows, and the runtime ones are causing some... issues:
We'll get there eventually I'm sure.
- Location
- The Land of Many Bregrets
- Pronouns
- They/Them
More Art Progress!
So as I said I should probably have done earlier, I started working on the tree room with the actual trees in it, from which the forest themed level might possibly benefit.
Here's a collection of WIP images whilst I was making the tree:
Leaves need work, both texture and shader, but I wanted to get them ingame just for testing purposes first. Since uv-mapping and texture painting the leaves of the entire tree would be insane, I broke it up into 'sections' I could repeat. It's greyscale deliberately; the idea is to use the 0-1 grayscale values to lerp two more leaf-like colours, so I can get variation in leaves across multiple tree instances. I'm also looking into using vertex colours to add more shading into the leaves (ie 'top bit lighter, lower bit darker') but the textures are kinda already handling that tbh. What's more important is to get some perlin noise vertex movement going on so that the leaves are moving the breeze. It feels like a pretty lifeless statue at the moment.
Once that's done I'll start sprinkling them about the outdoor areas outside of the playable space since, again, forest level.
One thing I do have to say is that this is the sort of workflow you could only do in blender. It involves modelling, uv-mapping and texture painting pretty much all at the same time; basic blockout shape, uv it, slap in rough guidelines in texture paint, model using guidelines, build up the model with repeating parts, iterate textures and uvs etc etc. Trying to do that in 3ds Max - which would involve repeatedly crunching uv unwrap modifiers, praying whatever model edits you made didn't break all of your work, bouncing into photoshop and trying to remember how a 2D image maps to a complex 3D shape, it would be... urgh. When I first started out modelling - over ten years ago now - I hated texturing. Now I can actually handpaint things by the power of blender and it's wonderful.
(perhaps the biggest proof the leaf mockout textures were done entirely in 3D painting is that the texture uvs are actually upside down. But hey, 3D paint, who cares!)
So as I said I should probably have done earlier, I started working on the tree room with the actual trees in it, from which the forest themed level might possibly benefit.
Here's a collection of WIP images whilst I was making the tree:
Leaves need work, both texture and shader, but I wanted to get them ingame just for testing purposes first. Since uv-mapping and texture painting the leaves of the entire tree would be insane, I broke it up into 'sections' I could repeat. It's greyscale deliberately; the idea is to use the 0-1 grayscale values to lerp two more leaf-like colours, so I can get variation in leaves across multiple tree instances. I'm also looking into using vertex colours to add more shading into the leaves (ie 'top bit lighter, lower bit darker') but the textures are kinda already handling that tbh. What's more important is to get some perlin noise vertex movement going on so that the leaves are moving the breeze. It feels like a pretty lifeless statue at the moment.
Once that's done I'll start sprinkling them about the outdoor areas outside of the playable space since, again, forest level.
One thing I do have to say is that this is the sort of workflow you could only do in blender. It involves modelling, uv-mapping and texture painting pretty much all at the same time; basic blockout shape, uv it, slap in rough guidelines in texture paint, model using guidelines, build up the model with repeating parts, iterate textures and uvs etc etc. Trying to do that in 3ds Max - which would involve repeatedly crunching uv unwrap modifiers, praying whatever model edits you made didn't break all of your work, bouncing into photoshop and trying to remember how a 2D image maps to a complex 3D shape, it would be... urgh. When I first started out modelling - over ten years ago now - I hated texturing. Now I can actually handpaint things by the power of blender and it's wonderful.
(perhaps the biggest proof the leaf mockout textures were done entirely in 3D painting is that the texture uvs are actually upside down. But hey, 3D paint, who cares!)
- Location
- 委員長の牢獄
Looking good!
My personal opinion is that you can get away with almost anything as long as the end product is both functional (a higher bar than one might think sometimes...) and has excellent aesthetic. There's a power to art-driven design that purely gameplay-functional design just can't compete with.
Or maybe that's what I feel because art is my weakest skill in terms of game dev
Still, if you can find the chance to introduce some environmental storytelling - past the hill of getting everything actually functional - I think that would be a good way to elevate what you have!
My personal opinion is that you can get away with almost anything as long as the end product is both functional (a higher bar than one might think sometimes...
) and has excellent aesthetic. There's a power to art-driven design that purely gameplay-functional design just can't compete with.Or maybe that's what I feel because art is my weakest skill in terms of game dev
Still, if you can find the chance to introduce some environmental storytelling - past the hill of getting everything actually functional - I think that would be a good way to elevate what you have!