The Longhouse

Mini LD 10: After School

2009-07-02T12:00:00.031+01:00

After School is a game I made for the tenth Mini Ludum Dare compo. Play it at yoyogames.com, download it here, and see the original compo post here.

design

I made this using GM7. I don't even have the pro edition. There's a pretty major bug for XP users, and a couple things I'm not happy about.

-it's not clear if you're stuck in an infinite loop - players could quit and miss the final message.
-the art sucks
-some of the writing isn't great
-the fonts have weird fuzz on there
-there's no sound
-the animations are corny
-it really sounds like I hate it, doesn't it?

But I don't hate it, not even a little, because this is the first game I've ever made and it was an education to persevere with it. I came up with the idea for a game you couldn't win because I thought it'd be insensitive and ignorant to make a game where you can just talk the abuser out of being violent - these ~~people~~ scumbags have already given in to their urge to hurt you, to shatter your confidence and make you feel ugly and threatened on every level, and all you can do is A) be a victim, or B) hit the escape key.

code

I used a couple things. GM7 works with objects (which are secretly classes), instances (which are secretly objects), and scripts (which are secretly functions). I have a dude called Kubrik; he's a persistant object that contains a very clumsy List of Shit To Do. How it works is, Kubrik runs through a loop with a value called "step". If step = 1, print the title card (this is a couple of dialogue boxes). The dialogue boxes are programmed to tell kubrik when they are done. If kubrik makes three of them, I have to tell him to wait for three of them to finish. This isn't automatic or sexy in any way.

Then I have an object called dialogue_response, which is an object that prints text and detects if the user moves the mouse onto that line. If so, it highlights it red, and if they click, it passes a value to kubrik, then kills itself and all other responses.

This value is Kubrik's new value for "step". So, right away, my program doesn't remember the result of previous decisions. It's also pretty limited to dialogue, characters walking on a flat plain, and stupid animations. I don't have many tricks up my sleeve at all.

And it has a bug where XP users can't highlight dialogue options.

life cycle

I started off wanting to make a game where you start as a baby, then become a kid, a teenager, and a young adult, at which point a certain thing happens. I won't say what the thing is, because it's libellously close to something that happened to a real person. Suffice it to say, it's the ugly pinnacle of everything "domestic violence" has to answer for.

The game also was too large, and not interesting enough. When I came up with the idea I did, I was pretty pleased at how concise it would be. That was on day 2, though, and I was having trouble just getting things to print on the damn screen. In the end, it took me four days to actually finish it, and a couple people had nice things to say. I emerged a wiser developer. Then the bugs starting raining from the sky.

Still, I'm tempted to release a slightly better version. Hmm. Any thoughts?

Ludum Dare 15: Caverns

2009-08-30T04:52:00.002+01:00

The theme is a little bland, really. It's like the theme being "fields" or "playground" - it's a place, not a thing, not a theme. With other places, like "mountains" or "valley", there's more than a place - there's a resolution, a complete object. Maybe you need to climb the mountain. Maybe you need to escape the valley, or find the secret at the centre of the valley.

I started in GIMP yesterday and sunk some chilled hours into spriting. I actually love spriting, it's really cool. I get daunted by the big tasks but it's no big deal, really, once you get into it.

It's not coding though, is it? Coding wise, I got to that blank page and hit this wall. My game is a tunnelling game where you need to collect resources and build underground complexes, and fight mad blob things that eat your dudes. The problem is this: how do I populate the entire grid with dirt, the occasional cavern, and stick a starting point on there? I guess I got scared and I was pretty tired and just wanted to make some dinner and spend an hour with my wife before collapsing into bed.

So I did that, and now it's 5am, and I have about a hundred and fifty unread LD48 blog posts. Fuck that. I also had a bunch of tweets, but they're only short, you know?

So I've got some cool ideas bubbling away here, but I think I know what to do: get a playable version as soon as possible. Like ditch the graphics and worrying about getting their hands to match up with the ladders. I can make a dot game. Hmm. I'm just not sure how to make negative shapes and have those be the objects, you know? Rather than defining all that dirt, define the absence of dirt. Yes...

Okay, Scratch That

2009-06-24T08:23:00.001+01:00

So the point of indie development is to develop little games quickly, right? I mean, not always, not really - look at Darwinia, for example. But it's one of the perks, you know? To work on a game for as little as an hour and then release it. Wow! Imagine doing that 8 times a day, 5 days a week? You'd be like increpare or some shit!

Anyway. I'm scrapping The Crossing for now. A couple reasons:

1) It's way over my head - I don't even know how to do some of this shit. I'll try actually making a proper game before I re-define the entire industry, maybe.

2) There's already a game or two called that.

3) I'm not enthused about it anymore.

So.... I'll probably wait to unveil things as I work on them. I intend to keep this blog as an open devblog that hides nothing - this means I'll make no promises, neither. Valve learned that one the hard way, I think. Poor Valve.

Where to go next? I'll probably stick to Game Maker 7 and attempt some Klik & Plays, some Ludum Dare things, and mess around with Valve's SDK, squeezing all that into my other life as a busy games journalist. I have a couple of good ideas, but just playing some indie games and remarking at how simple a great game can be, maybe I don't need to procedurally generate everything. Even though I really want to, BTN.

Paraphrasing is hard

2009-05-19T12:00:00.004+01:00

Edit: computers? Understanding emotions? FROM TEXT? WTF.

It's easy to come up with weird sentences like: "The Captain is destroyed" if you just use a thesaurus replacer. Hmm.

I'm thinking a lot about the way we construct sentences and it almost seems like you can't just replace words, you have to randomise the structure as well. Then you get sentences like:

"Happy today, I am." Which... nobody would say. Hmm.

I can write simple things (my name is #1. i have a #2.) and switch them randomly with more complex things (hey, have you seen my #2? i'm #1, by the way, hi.), but it's tricky to anticipate everything I want to write.

Also, it's been hard to really get into a serious coding session lately, because my writing has been picking up speed, and I'm always working. I need to take a long holiday from work and set up some posts to schedule on my main blog.

Whelp, that's the weekly update! I'm gonna try and update this on my progress every Tuesday. It'd be fun to get my engine to work on the updates :D

Hello #player#. Did you see that #event#?

2009-05-09T09:00:00.001+01:00

Very interested in doing something with procedural text. My efforts with The Crossing might actually require a lot of chasing something down, but pipe-dreams are fated to take fucking ages, so there you go. Game Maker borked, somehow, so while I fix that I'm thinking about a fantasy JRPG setting; you know the kind, the town is full of people who only say one stupid thing over and over again? My proof of concept will be to get them saying the same stupid relevant, procedurally generated thing over and over again.

As a tertiary goal, I must not make this AI-Complete.

Also, I'd like to say that I've been asked to make a game about working in a gas station, where you get points for adding snacks onto their fuel sales, and spend the points on guns. So like minataur china shop, but with guns and people.

I miss the sensibilities of the original Doom. Quake Live helps.

Following Procedure

2009-04-13T09:00:00.000+01:00

I've been mostly playing with the ship layouts. Navigating is easier if the ship radiates out from a central elevator spoke.

Creating a procedural ship would be a satisfying challenge, and I'm almost done running through the logic. I don't want to make a game that isn't procedural in some way, I've decided.

I don't know if I still want to do the Monster thing; being the bad guy is fun in Overlord and GTA, but this isn't a game about torture. Maybe I could do something truly unhinged with the monster, make you a sad, emotional leech that just wants a (potentially fatal) hug. Still, that's a problem for a functional game.

For now, a series of rooms with people living in them would be great. Maybe I should be making a fish tank of people that responds to stimulus? Maybe, for now, I should just let you read their thoughts?

These questions will be answered in the next update, hopefully this year.

I can Talk the Talk

2009-02-10T12:00:00.000Z

Okay, a short one. I'm at the stage where I'm getting an idea of the grand, impossible task I've set myself. I'm equal to it, though, because all those years spent swearing at games and claiming to be capable of doing better must be avenged.

So I'm working on the dialogue. I have a system where my director object ("cKubrik" :D) creates scenes, objects parented to an object that runs lines of dialogue; all the children have little screenplays in them, and they all print the lines above actors for the specified length of time, and signal the actors to change their sprites, etc. It's almost left the design phase. Now I'm wondering how the hell to make them walk about; the actors move about in an area the player can't generally enter, and they'll need to roam about the map on a complex schedule. I want to code this myself at the risk of re-inventing the wheel, and it'll just take some thinking to break it open.

Finally, I'm probably not going to use the bubble ship. I might not use the invisible monster either, although... well, I don't know. My plan is to wait until I have disposable income and grab the pro version of GM7, even just for the transparency. Spriting is getting better, I feel like I'm improving.

I'll update the progress thing with more realistic goals soon, maybe before the week is out? Not sure. This isn't top priority, but nothing else is really going on right now, so away I shall work.

Thanks for tuning in, folks!

Some core decisions

2009-02-01T09:00:00.001Z

Wonderfl is seriously awesome.

So, I'm deciding between many methods here. I've got the following dichotomies:

Platform vs Adventure
On the one hand, I want left to send you left, right to send you right, and when you lift those keys, you should stop. I would like, which is to say that it isn't indispensable, for the act of killing to be a tense skill challenge. Pointing and clicking is right out.
On the other hand, you can investigate items and navigate menus. It almost sounds like an adventure game, like Monkey Island in Space where you play as LeChuck, but I want it to remain solid, I want to see a little health bar as you drain the life from a crewmember. Hmm.

Invincible vs Vulnerable
While combat should be tense, I don't want players frustrated by the skill of an npc. I also don't want to make a game out of "run at people while they shoot you with very advanced weapons". But can it be done? If you're never in any danger and all they can do is run, won't you just plod along until you corner them? Where is the urgency?

The fact is, in the Crossing, you don't live and die by the sword. There is no "Game Over" screen when you fail; you know your goal, and you know your failure. It's a little sand boxy in that way. But adding player health could add some tension. Maybe a big health bar could appear when you're shot, and the miniscule portion that vanishes as you get hit just feeds your experience of being an intimidating monster. Or maybe npcs could pull out a firearm and shoot you to no effect.

Scripted vs Dynamic
I could give the NPCs a loose daily schedule, something like an AI package for work, a package for recreation, meals, and sleep. I could let little disputes and romances boil up and whirl off like dust bunnies. This is close to my original idea for the game, but it's far-fetched in its full glory.

What I could also do is tightly script the whole thing, and branch off many paths. I don't want to cut corners, so this would be...

4 crew members, each with an important job and a distinct personality.
1 dies every day for 4 days; this is the game.
Every death affects a person differently at every stage, so in effect, there are 24 endings. Start out by killing x's loved one, and he might not even bother to call for help. Man, this game is sick. And not in that annoying way that means "really good", I mean like, if you play this game, (thankyou and) you should be locked up.

Getting back on track, scripting it will be tedious and only a little fun. Writing AI is always fun, but it's easier this way to forget something. This is my main problem right now, but I do have one more.

Goodfellow vs Moloch
You play as a monster. You aren't a monster in real life though (ARE YOU?!), and you don't want to kill people and hide the bodies and drink their blood (DO YOU?!), so how do I motivate the player to play the game? Make it necessary, and give them a bastard to kill first. When you've killed, you get stronger, and people get tense and say horrible things. There's a fine line emerging here between Righteous Fury and Power Overwhelming; if you're on a ship full of bastards, even a normal person might just accidentally open an airlock. If you're basically toying with helpless mortals, you'll get bored. Nobody knows why this hasn't happened with GTA yet. Maybe because it's really really fun.

Okay, so maybe this isn't a choice I have to make. Maybe there's merit to both sides. Maybe I'll have a few bastard qualities for everyone, and maybe I'll put in a "lift crew member ten feet up and drain them dry during Seinfeld in the rec room" mode.

Free vs Modestly Priced
Two things can happen. Number one, next month this blog is covered in dust. Number two, in two years I release an awesome game beyond any of my current expectations. For free. Like an ass. I want to make some money on it just as a matter of principle, so I can cite it as experience. But I don't want to try to sell something worthless; there are a lot of excellent designers, and I'm sure lots of people knock out vaguely interesting tripe in game maker. This is a question for the beta.

Where does the tile fly?

2009-01-29T09:00:00.005Z

Blambot is an excellent source of fonts. I'm trying some out for the various dialogue, thinking the creature might use a scary font, etc. BGHQ is good for downloading old megaman screens and examining them for artistic gambits. I suck at art, so that helps.

Reiner's Tilesets is a source I mean to investigate more thoroughly, because it looks full of interesting stuff, but it appears to have only isometric tiles and so on. Oh well.

I've made some decisions about the structure of the game. I have four Crew for now, and they're going to react to eachother in different ways because of their personalities. The design problem is that, in a game as open as this where every decision is open to you, I want to push the player so they aren't just totally unmotivated. So the decision I've made is that they're set in stone; you can replay the game... 24 times, say, for the major differences. This is a game where I plan to get the graphical stuff over and done with first, or last, and spend a lot of my time scripting (skippable) cutscenes.

I call them cutscenes, but the player will be able to interrupt them. If this is done before, for example, certain beans are spilled, it will change the outcome of events slightly. I want to try and break open conversation for dynamic scripting, the same way numbers have been broken open. Welding together sentences could get messy, but I could queue up paragraphs, etc. It's going to be hard to do it, but someone's got to. We do it in our heads every day, predicting what people will say, how they act. The algorithm is there already.

The ship is an unusual layout; I thought of this big reactor powering the drives and having the crew quarters sealed at the front for insulation. The reactor and the engineering part would be really cold. Have a look at this concept art:

The source image is layered with all the maps and so on, but it's only really that nub at the front and back you'll be playing in. The big orb is the reactor. I was thinking about an enormous cargo hold stretching down from the front section like a pistol clip, but that might screw with the plausibility of the game (any of those crates could be full of something the monster could eat, growing to full size and terror within a day).

The Crossing in early development

2009-01-27T17:00:00.000Z

An adventure game set in deep space. As a powerful, invisible monster, you must feed. But you're no brute; you understand your prey, even pity them. As your energy depletes, you risk returning to that void from whence you came, and you dare not contemplate that; but who should you kill? The arrogant Captain with the true heart of a hero? The introspective engineer that broods with a secret urge? The frustrated communications officer who just wants some fun before she's too old to have any? The Microbiologist who can't, for the life of her, find those tissue samples she left in the incubator?

Killing sates your hunger for a day, but causes you to sleep - your nest is in a safe corner of the ship, but there are only so many days before the Crossing is over. How will it end? How can you best influence events?

The Crossing

Currently building with Game Maker 7. This blog will track the development process, share code examples, and generally act as an inspiration (or warning) to wanabe programmers everywhere. Well, the ones who can speak english and manage to find this blog.

Current progress tracked on the sidebar.