ADEPT - playtest build available!

[Jwatt]

https://www.youtube.com/watch?v=JXDIF7Yxj5I

Recorded a gameplay video! It's basically me messing with the editor and playing the demo level. Featuring hand-picked Adlib tunes that wouldn't be out of place in the actual game.

[airstruck]

Just trying this out now. Some random thoughts in no particular order:

- I like the font on the help screen, except for 1 or 2 things. I keep reading the "g" as a "9" because what should be the descender sits on the baseline. I think an uppercase "G" might look better. I'm on the fence about the comma, I don't love it but I'm not sure how to improve it.

- I feel like escape should close the help screen, and escape on the main screen should prompt to "really quit" (and escape at the "really quit" prompt should close the prompt, returning to the main screen). Attempting to close the window through the window manager should possibly also be intercepted by the "really quit" prompt. I know this is an early version, just something to think about for later.

- I don't understand the meaning of "inf" under the cursor. I think I'd prefer the preview of the block I'm placing to be right inside the cursor (maybe faded a bit or something), since I can't place blocks where other blocks already are anyway. I keep thinking I'm placing things where the block preview shows up.

- The player character should probably always draw in front of things, esp. ladders.

- Sliding down ladders when not pressing anything feels weird.

- Falling ladders means there can't be a ladder suspended above the floor that you need to jump for.

- I like that you can have more than one guy in play, could lead to some interesting puzzles. Maybe you could hit a key to switch which guy you control, and you can set how many of them need to escape to win a level.

- After messing with this for a bit, I suddenly want to play Lode Runner.

Anyway it's an interesting concept and has potential, keep it up!

[Jwatt]

Wow, that is some thoughtful feedback. I'm feeling inspired to write about the design anyway, so I'll be giving some wordy answers.

- I like the font on the help screen, except for 1 or 2 things. I keep reading the "g" as a "9" because what should be the descender sits on the baseline. I think an uppercase "G" might look better. I'm on the fence about the comma, I don't love it but I'm not sure how to improve it.

I can see your point. I might have to edit the font file myself, because I'm easily annoyed by these things too. Perhaps the comma should be like this:

███
███
██

- I feel like escape should close the help screen, and escape on the main screen should prompt to "really quit" (and escape at the "really quit" prompt should close the prompt, returning to the main screen). Attempting to close the window through the window manager should possibly also be intercepted by the "really quit" prompt. I know this is an early version, just something to think about for later.

Oh yeah, absolutely. There really isn't a concept of "game state" as such, after all there's only one scene. Heck, there's not even a distinction between playing a level and editing it. But when I have some sort of scene structure, any immediate quit buttons are straight out.

- I don't understand the meaning of "inf" under the cursor.

Don't stress it, it's just a placeholder saying you have [inf]inite tiles. In the future with limited tiles, it'll be some number representing how many of those tiles you currently have to build with.

I think I'd prefer the preview of the block I'm placing to be right inside the cursor (maybe faded a bit or something), since I can't place blocks where other blocks already are anyway. I keep thinking I'm placing things where the block preview shows up.

Yeah, designing the cursor is quite a challenge. Especially considering your selected block is going to have more sub-options, or "modifiers".

You know how the LED screen shows you some descriptions of a block's behaviour when you use the scrollwheel? That's what I call modifiers, and the plan is that you can combine basically any unlocked modifiers with any tiles.


For example, this would be a spring that does bounce you as usual, but it also travels counter-clockwise across the map.

I'm sure you realize the implications - spikes flying at you across the map, coins that trigger some nasty trap event, fake exits...

- The player character should probably always draw in front of things, esp. ladders.

It should, and I thought it did since day one. I had to go and check, and clearly this is not the case anymore. Must've been a typo when I was reordering layers, and somehow I haven't noticed after that.

- Sliding down ladders when not pressing anything feels weird.

The way ladders work is that you're currently either holding onto them (so pressing a key) or not. The "sliding" is just gravity - the weirdness might come from the player not having a pose for holding onto a ladder.

Based on the way ladders work in most other games, you'd expect to have an additional state describing whether you've become attached to the ladder or not. It's just that this makes ladders clumsy to use in Adept. Ladders aren't just a synonym for stairs/elevators in this game: they're physical objects that can be destroyed, dropped into pits, pushed around, used to block enemies, break fragile objects, activate pressure plates... The controls (and many upcoming obstacles) lend badly to state changes in the player's controls.

This is also why you have to hold left/right if you want to ride on a horizontally moving platform - maybe in some designs, the player would be intended to hop off moving platforms constantly, and the player going into a "riding" state would be a hindrance. I want to make as few assumptions about a tile's use cases as possible.

- Falling ladders means there can't be a ladder suspended above the floor that you need to jump for.

Yes, that's an unnecessary limitation. I was already thinking of adding "support beams" which don't collide with the player, but do collide with other physical objects (and ladders specifically). You could place one of those in the air and stack ladders on top.

- I like that you can have more than one guy in play, could lead to some interesting puzzles. Maybe you could hit a key to switch which guy you control, and you can set how many of them need to escape to win a level.

I'm glad somebody noticed! Yes, I've made sure the player has no special privileges. It's like any other tile: neither specified to be unique nor required to exist at all. I believe that in decently well-planned code, any amount of any object should be allowed in the scene by default, and it should take special effort to add limitations that say otherwise. The entity-component system used here is great for emergent use cases.

Having multiple players isn't just a curiosity either, it does open gameplay possibilities like what you said.

- After messing with this for a bit, I suddenly want to play Lode Runner.

Ohh, you're very good! Besides the general retro aesthetic, Adept does borrow quite a bit from games like Lode Runner, Jetpack, and N. I've always viewed those three in particular as part of the same continuum, and I've long wanted to make my own spiritual successor to them.

The thing that brings these games together in my mind is that they consist of simple pieces which can come together in very varied ways. There isn't really any object that's a singleton, everything has potential for clever interaction with other objects. No wonder their level editors have added so much longevity to all of them. Each one of these has also pushed forward the way we share level data.

But none of them has yet gone off the deep end and made level design itself be the core of the gameplay. In Adept, one-upping each other with even more creative designs is the overarching game to participate in. Features like the dynamic ranking system and tile limitations serve to support that.

That's the next logical step, and the time is ripe: we have the potential of modern tools like Löve, and the indie spirit of repurposing classic ideas is in full bloom.

[Jwatt]

For today's update, I think now is a good time to flesh out some trendlines for future development.

tl;dr: me babbling about gaemz but no gaemz. If all this talk doesn't interest you, skip this post.

______________________________________

As it is now, the architecture of the editor is mostly done. At least when it comes to adding new types of tiles. Tiles are the core content in this game, and your progression depends on how many you have unlocked.

Since I think dividing stuff into themes is inspiring, there will be less than 10 quite different tilesets. Each tileset also has the gameplay purpose that it constitutes one ranking system: you can be the reigning champion of the cavern world, while owning the rank 3 level in the jungle world, for example.

Progression

As you start the game, you have no tiles available to use in the editor. That makes sense, because how would you design a level without knowing what each tile does?

By playing through the different worlds, you'll encounter new types of tiles that other users have placed in their levels. Once you complete a level, you're deemed to have the minimum understanding of every individual tile type in it, and so you unlock all the new tile types encountered in that level. Think of this as a Pokedex for tiles that you expand by playing through others' creations.

Once you've unlocked enough of a world's tileset to make a level of your own, you're free to submit one there. Freshly submitted levels start out at the spot which the player has reached in the world's ranking: if you've played through 4 levels, your level will start out at the 5th lowest rank, pushing the lower ones further down.

Remember that level rank determines the order that new players in the world encounter levels: everyone starts at the bottom, where the easiest levels would ideally fall down, due to them having the lowest ratio of deaths per attempts. (If you leave a world and come back when people have submitted more levels, you'll continue from the same index starting from the bottom, there'll just be more ranks ahead of you.)

Of course, you'll have to win your own level before submitting. And you don't have to create everything in one go, you can come back to fine-tune locally saved levels before submitting them.

Fully unlocking all the tiles in a tileset opens up new worlds with their own respective levels and rankings. You can still try to finish all levels in a cleared world if you want to maintain a high player rank (or just stay in touch with how the world's meta evolves).

Most worlds have some limitations, typically only allowing the basic tiles in addition to the world's own tileset (to prevent sequence breaking for new players), or clear-cut restrictions like "max. 100 tiles per level".

Worlds

Here's some worlds that I definitely want to include. I plan to have a handful of very distinct worlds, rather than a whole bunch that only really differ visually.

• The cavern world is the place to start, like a tutorial world. It contains the very essentials required to make a playable level, basically the subset currently in the game: solid blocks, the player, exit, collectibles, spikes... Once you've unlocked these, most worlds allow you to combine them with other tilesets as well, so there doesn't need to be like a different exit door tile for each world. (But there can be, for decorative purposes if nothing else. I'd like to see different player sprites in different worlds, for example.)
  
  This cavern tileset also has a slight industrial subtheme, with unbreakable steel blocks, drills, mechanical moving blocks and so on. Due to a low amount of tiles allowed per level, you're bound to see lots of negative space, as expected with a cavern theme.
• The jungle world. This place has many traps that are triggered by line-of-sight or other means. Spears that extend until they hit a solid block, pressure plates that cause boulders to come rumbling down, foreground decorations like vines that can cover up otherwise obvious traps. Levels revolve around risk vs. reward, puzzles, and trap deduction.
• A high-tech world with gadgets, neon lights and such. Here you're introduced to one-way tiles, conveyor belts, magnet shoes that allow you to defy gravity and walk around certain surfaces, lasers with a blockable line-of-sight mechanism, timed electrified floors, quite a bunch of animated things that test your reactions.
• A candy world. The levels in this place revolve around the use of a bouncy ball companion - you can push it around to block enemies, walk on it to stay safe from hazards, bounce on it to reach higher ground, and so on. The ball can even collect coins for you (which are reskinned as candy here). It also serves as the level exit to ensure that it gets used.
• The crystal world, an odd place where events are introduced - touch this, and something is triggered somewhere. There are only a handful of different tiles in this tileset, so expect to see copious amounts of events used creatively. After all, events are modifiers unlocked with the crystal blocks, so there's an infinite amount of them. Levels might have completely rearranged themselves through the use of events and moving elements by the time you're done with them. Or you might have rhythm levels with timed enemies triggering events repeatedly. And with the use of triggered moving blocks to describe state, someone somewhere is bound to make a Turing-complete level.
• The final stop is The Galaxy. Or perhaps, that's where the game truly begins. In The Galaxy, everyone's free to use tiles from any tilesets. No restrictions on tiles per level, either. And everyone here has played through all the other worlds, so you can assume that both players and submitters know all the tricks. It's a no-holds-barred global battlefield where only the most clever players and creative designers can reach the top ranks.
  
  The only catch is that in The Galaxy, you no longer gain new tiles by playing through levels. That way, you still have a reason to keep in touch with the lesser worlds, collecting loot to prepare your ultimate creation for The Galaxy.

[Jwatt]

A bit of foray into music before getting into networking.

So, FM is my jam, and that's what we're gonna use here. I happen to make some music now and then, and I've always had original creations as my game soundtracks, even making some tracks for others' games. I quite enjoy thinking about what sort of moods would fit different scenes, and I find audio to be sadly an often unappreciated part of games.

I did carefully pick some fitting DOS game music for the gameplay video for a reason - check it out if you want to hear what sort of aesthetic I'm going for.

My setup for this game is Renoise with JuceOPLVSTi. Renoise is a tracker-like music workstation with all the modern bells and whistles. Juce is a plugin which is basically an emulator for the Adlib soundcard, but is much more convenient than vintage Adlib software and contains tons of preset instruments as well.



Srsly, that's just one of 4 huge folders.

[airstruck]

Since I think dividing stuff into themes is inspiring, there will be less than 10 quite different tilesets. Each tileset also has the gameplay purpose that it constitutes one ranking system: you can be the reigning champion of the cavern world, while owning the rank 3 level in the jungle world, for example.

Love it.

By playing through the different worlds, you'll encounter new types of tiles that other users have placed in their levels. Once you complete a level, you're deemed to have the minimum understanding of every individual tile type in it, and so you unlock all the new tile types encountered in that level. Think of this as a Pokedex for tiles that you expand by playing through others' creations.

I guess you'll need to seed it with some levels so players can get those tiles from somewhere initially (chicken/egg problem). I like the idea of unlocking new tiles as you through a world, but I'm not sure I like being introduced to the tiles in an arbitrary order.

What about this? Instead of building levels on a totally blank canvas, there's already some immovable stuff there. Not much, maybe just flat ground, or a couple of hills, or a mostly empty cavern with one or two big platforms, stuff like that. These would be the levels within the worlds, they'd have their own names and leaderboards and you'd play through them in order.

I'll call what the players are building "bases" instead of levels. Once you beat, say, 3 different bases on the first level, you can build your own base on that level, using the subset of world tiles allowed on that level. You'll have to build a base before playing the next level (will help generate content).

Now, after you build your base and beat the next level, you'll unlock more tiles, and you can build a new base on that level. Maybe you can also go back and add these new tiles to your base on the first level. This would put your first level base in a tougher class. After beating a world, you could play back through the levels with these tougher bases.

Anyway, just throwing that out there.

Oh, and enemies? I'd like to be able to place a few enemies to guard my base, not just traps!

[Jwatt]

I guess you'll need to seed it with some levels so players can get those tiles from somewhere initially (chicken/egg problem).

That's the plan. Initially the game would have a handful of seed levels created by me and some of the testers. Together, these levels have to use every tile in a tileset at least once, of course. After that it's exponential growth, because tiles get copied, and a new level can contain up to all of the type X tiles ever placed, and the next player's level can contain all that as well, etc.

I like the idea of unlocking new tiles as you through a world, but I'm not sure I like being introduced to the tiles in an arbitrary order.

I doubt the order is all that arbitrary. There's probably going to be some manner of sorting happening naturally, because some tiles are just easier to deal with.

But I do have a concern about whether some of the more advanced tiles will be used too rarely. Like, imagine a trap tile with some complicated behaviour that always gets a few kills out of unprepared players. Levels using that for cheap kills would rise higher in the ranks, becoming late-stage curiosities. The worst case is that people would have to play through potentially hundreds of levels before even seeing a tile like that once. So you really don't want this crowdsourced sorting algorithm to be too good, heh.

Most likely deaths per attempts shouldn't the only factor in defining a level's rank. A rank isn't only supposed to measure raw difficulty, it's also about how effectively you've used what you have, and a general gist of player involvement required (length or complexity or whatever you want to call it). Earlier levels aren't just easier, they're also going to be generally simpler in design than late-stage levels. Hard to say exactly which variables should count without having an actual ranking system yet, but my first guess is:



Edit: ...Though you'll also want each individual term to have some emphasis factor, which gets fine-tuned through playtesting. So it can't be pure multiplication, or you'd only end up scaling the whole result. Also, tile count needs not factor into rank, because it's not a quality that changes as people play the level. It would just lead to people spamming tiles anyway.

So here's a better formula that produces much more reasonable ranks for various test inputs:

Yeah, I'm totally geeking out right now.

What about this? Instead of building levels on a totally blank canvas, there's already some immovable stuff there. Not much, maybe just flat ground, or a couple of hills, or a mostly empty cavern with one or two big platforms, stuff like that. These would be the levels within the worlds, they'd have their own names and leaderboards and you'd play through them in order.

I'll call what the players are building "bases" instead of levels. Once you beat, say, 3 different bases on the first level, you can build your own base on that level, using the subset of world tiles allowed on that level. You'll have to build a base before playing the next level (will help generate content).

Now, after you build your base and beat the next level, you'll unlock more tiles, and you can build a new base on that level. Maybe you can also go back and add these new tiles to your base on the first level. This would put your first level base in a tougher class. After beating a world, you could play back through the levels with these tougher bases.

I'm liking this, it's a more refined approach to this whole social experiment thing. Instead of just "do whatever", you'd have a bit of a push, some slight guidelines to challenge people. Like giving musicians a scale and a theme and letting them improvise on that. It's not at all far-fetched for what the game is about.

The Galaxy is obviously a place where anything goes, but I can see the regular worlds benefiting from templates like that to create a better framework for competition.

I'm going to give this some more thought, thanks for the useful feedback again.

Oh, and enemies? I'd like to be able to place a few enemies to guard my base, not just traps!

Enemies are definitely coming!

[CanadianGamer]

A lot of fun.