I’m tired of failing to make games, so I’m in. It’s not like I can’t make games in theory (I have degrees and shit), but as projects go on they tend to die, and their corpses litter my harddrive. So it’s time to make a game. An actual game, and not a loose collection of code that that really really wants to be a game one day but can’t sustain my enthusiasm long enough to make it. So I think Ludum Dare is pretty much perfect to kick things off.

I’ll be using HaXe with haxepunk for the programming. I’m not very familiar with either of them, to be honest (been playing with them for the last week), but I’ve done a lot of AS3+Flixel stuff, so maybe it’ll work out.

I guess Gimp for graphics, sfxr/Audacity for sound effects, and maybe Otomata for music. I doubt I’ll have any time to be ambitous on the sound front, and it’s not like I have any instruments lying around – but we’ll see how important it becomes.

I also live about 20 minutes by bicycle from the Tokyo meetup in Kichijoji, so that should make for good motivating. Nothing quite like desperately coding in the middle of the night with other people.

Here’s the result of me playing around in Haxe/Haxepunk for the last few days. Not very much, but by dropping it here I guess I can save a bit of time, so: http://www.mediafire.com/?h61lj5y3n3kxf6s

I wouldn’t really recommend that anyone else tries to use this code for anything – I don’t know what I’m doing and all it amounts to is player input and falling anyway, but there’s no point in writing it again if I need it so…

Anyway, off to bed. When I wake up again it should be pretty much time to start. For once, the timezone thing works out well.

So I’ve started making a game. The rough idea is that you’re a witch who has turned the princess into a goat and taken her place. The goal is something like kill the rest of the royal family while foiling the goat-princess’s attempts to stop you. Or something. I’m not yet 100% clear on the gameplay.

I have minimal faith in this idea, but I’m rolling with it because I don’t see another one coming along any time soon.

Anyway, off to Kichijouji now for the meetup. Hopefully good ideas will come in transit.

Alright, it’s now 11pm, meaning that Ludum Dare started 12 hours ago. How are things?

Kind of bad. I don’t understand why game companies would try and make programmers work for 12 hours a day. Beyond 8 hours the productivity drop was pretty huge.

I think the chances of finishing what I had in mind are pretty small, and I don’t consider what I had in mind all that ambitious. I’ve just finished (‘finished’, in that it fully worked for the first time but I bet there’s all kinds of problems going on there and more to do if I can get around to expanding it) the attack system, so basically you can walk up to the King and stab him, and he dies. Hopefully it should be easy to expand that to having an enemy walk up and stab you, but my code architecture is turning into a huge clusterfuck.

Plan for tonight is to get up to the guards walking around and stabbing you part. I think then there will be technically enough mechanics to call it a game, after level design. After that’s done I’ll go home and sleep for a bit. Or I might before that, depending on how derpy my programming gets as the night goes on, but I would rather spend tomorrow implementing other things. Also, it would be good to get the art out of the way, because making art when I’m feeling alive is a bit of a waste.

Right now though, I think I’ll go for a walk. The freezing miserable cold should wake me up, and maybe find some food down there. Is kind of convenient being in the middle of Kichijoji, I guess.

I just want to post something to prove that a game exists. And then I think fuckit, going home to sleep.

So here it is!

… yah, not so exciting.

I think I’m going to have to throw away half the gameplay I had planned to get it done in time. Very probably, you will just walk around the castle stabbing people while getting chased by the guards. I’m not sure that any plot will be implemented, so it might just end up being a medieval knife massacre simulation.

I shouldn’t have disrupted my sleep schedule. Can’t focus, can’t keep a grip on the big picture of what my code is trying to do anymore and how the bits connect together. Have been bogged down trying to fix the AI system for the past several hours, so now a bunch of stuff that used to work doesn’t. If I can get it doing what it was doing before, that might be good enough.

Anyway, lesson learnt, I think. The second day is the one you don’t sleep on. There really isn’t any point to not sleeping on the first anyway.

I might throw it in here. Nothing went well today, and my increasingly feeble attempts at poking this code into submission aren’t going to achieve anything.  Dead tired and dinner nearly killed me (holy shit who can eat anything that spicy?).

I might just go home, go to bed, and tomorrow or the next day rip apart my code and turn it into a different game keeping the good stuff and getting rid of the gunk that’s making it useless now. The general idea behind it isn’t so bad, but the stuff gluing it together is rubbish and it’s not clear when and how anything is communicating with anything else.

At least my game idea wasn’t much good, so no guilt in chucking that. Might make my end of the world idea instead.

Guess a post-mortem is in order.

So I didn’t finish. This was my first LD, and in general I don’t have a good track record of finishing things. I was using tools that I’m not very familiar with, but that wasn’t the main problem. My idea wasn’t, perhaps, as self-contained and unambitious as it should have been, but I should have been able to get something playable out of it given where things were when I gave up.

I’m going to drop most of the blame for this one on a lack of sleep. There’s really no good reason to stay up all night on the first day, but I worked from 11am-4:45am. It was pretty productive work – even towards the end, while things had slowed considerably, progress was still plodding along.

I got to bed at 5:30 and woke up around 11:30. Made a todo list and then failed to get through a quarter of it. The way the code fit together was fine in my head when I went to sleep, but the next day getting a grip on how the various things related and communicated to each other was far too much to cope with after not enough sleep on a disrupted sleep schedule.

And, of course, my game architecture deserves some of the blame for that. Lots of the blame actually, but it’s not so bad. I tried something I hadn’t done much before (components) – it was a good learning experience, certainly. The main idea was pretty good, but the glue holding it all together wasn’t. I’m planning to start a new game and scavenge this game for code – it should hopefully amount to something more stable and useful.

Anyway, that’s that. No grand reason for failure, just far too much enthusiasm for being hardcore on the first day. Being hardcore is for the day it’s due and not before – anything before that is wasted effort.

Tags: post-mortem

I’m in.

Tools are haxe/haxepunk, Tiled for the map, and Gimp for graphics. Nothing too fancy, I think.

I’m hacking the code from my other project which was languishing in the depths of apathy into something that can be used for this. So far the test map is displaying. Shouldn’t take much to get the rest of the basics up and running. From there I’m keeping my hopes and expectations low..

So I’m in. May this time go better than last time (all you can really ask for, really).

Code: Haxe/Haxepunk

Sound: Probably nothing, but sfxr or some other procedural sound thingy if I get around to it.

Art: gimp, I guess.

And base code, which will be the ruins of the last three games I was working on. Should be able to pull something from the rubble there (although I’m still not sure the way I’ve written it makes good sense … but whatever. Guess it works) http://www.mediafire.com/?yzp80icqk7a4zx4

I guess I should post-mortem my game now that the ratings are out.

http://www.ludumdare.com/compo/ludum-dare-27/?action=preview&uid=7956

Well, basically, my game wasn’t very good, but wasn’t particularly bad either. In fact, it got a 2.50 overall score, so I guess the word that comes to mind is ‘average’ or ‘mediocre’. I can deal with this.

What went well:

Haxepunk and my dodgy engine that’s slowly forming over the top of it:

There are some parts of my code that aren’t sensible and it needs a lot of tuning, but I can make things fast and slot in code from other games and it just works without lots of copying and pasting and awfulness. In general the programming went very smoothly – sure, there were a few instances of getting tripped up for an hour or so while some stupid bug caused by my own stupidity wouldn’t die, but generally speaking the code flowed freely.

The music:

I certainly didn’t expect to put this in the ‘things that went well’ section, and I feel kind-of bad that people liked the music because I spent about a minute in Otomata making it. It literally took no effort on my part and it was the very last thing I did before submitting and crashing.

What went poorly:

Time management:

I had flat-zero ideas for the theme and if I hadn’t woken up and found it a rainy miserable day I probably would have given up. I started a full 24 hours late. I also spent a few hours at local festivals that were going on in the neighbourhood, though I don’t think that had much effect. I started work at 10am and submitted around 4am the next morning (deadline being 10am my time) – I probably should have used the remaining time to polish but it was playable and I didn’t care anymore.

The ‘features’:

I tweeted around 3am: ‘Alright, there are no more bugs – only features!’.

Which is all well and good, but unfortunately no one else appreciated the ‘feature’ of the collision detection breaking 30 seconds in. It’s not as broken as most people thought, I think – you can still kill them with some effort and getting them to cross the hand a few times, but … yeah, it’s a bug that I decided not to fix because it was really late and I knew it would be really hard to fix. Plus it adds some challenge to the game so you can’t just sit in front of the hand … which is a stupid excuse. As my sister commented: ‘It’s really hard because at some point there’s so many guys breaking the rules that you can’t keep up’.

The movement bugs were well received though. There were actually no movement patterns besides ‘chase the player’, but when I changed a few values it appeared there were anyway. This is good, because that was part of the original plan that I didn’t get around to implementing.

The graphics:

I can’t believe I went with a shit-green-and-grey colour scheme. Anything would have been better, and it would have taken ten seconds with the paint bucket tool.

Anyway, enough navelgazing. It’s a miracle I released anything, so I’m not complaining in any case.

(working title, I dunno)

Alright, I’m done for tonight, I think. I’ve been plodding along all day and don’t feel I’ve gotten as far as I need to be. Energy levels just dropped somewhere along the line. Hopefully the looming-deadline effect will kick in tomorrow.

The concept of the game is that you’re Jesus, and one of your apostles has rocked up with 5000 people and expects you pull some kind of fucking miracle or something and feed them all. Wanker. Anyway, he gives you one loaf of bread and a fish and tells you the party is on in the evening.

You could tell him to go fuck himself, but you’re Jesus and you have a reputation to protect, so somehow you need to multiply your supply of bread and fish by a lot. Unfortunately, miracles don’t just happen on the drop of a hat. You’re doing it the hard way.

And so you’re meant to trade with other people for things they want until you’ve multiplied your bread and fish to the required amount.

This idea sounded less ambitious in the morning, but oh well.

With all apologies to Jesus, here’s Jesus:

Time for an update.

Progress has been bad, but not bad enough for me to give up. Which is unfortunate. I was hoping to sleep.

It took far longer than it should have, but the trading system works and as far as I can tell doesn’t break. The code running it is disgusting. Absolutely horrid. I hope it doesn’t break because I can barely remember how it works. It’s not even doing anything particularly complex – just the way it’s being done is stupid.

But anyway, I’ve moved on from that, and got on to the business of making the actual game. Well, so far just some terrible art, but turning that into actual game should hopefully be easy from here.

I leave you with the cast so far:

No, I’m not making a zombie game. I’m merely very bad at this.

I’m in.

The goal is to submit something playable, all else be damned. I will not give in to the temptations of ridiculous scope!

Here’s my halfarsed base code: http://www.mediafire.com/download/k9s1f20knleaf9d/src%282%29.zip

Nothing very exciting there though. I had nicer and much more code before I reformatted my hard-drive, but because I’m retarded it is now gone. But oh well.

Tools:

Haxe/OpenFL

Flashdevelop

Gimp and/or Inkscape

Idunno about sound. I’ll try work out something. Or not, depends.

Time to ponder out-loud.

I spent most of the day implementing pretty basic stuff that should have been in my base code already (input, movement, camera, collisions, gravity…). As I’ve been pottering along doing this my enthusiasm for my idea has slowly been draining away.
But that’s okay, I think. Tomorrow is when shit will solidify into something real. Probably.

The problem with the current idea (basically: rappelling/falling down into a place with messed up physical properties that kill you and using various tools to see how far you can get) is basically the same as the problem with my previous LD – it’s relying on me having enough time and energy and life in me after implementing all the mechanics and such to come up with a bunch of content for the actual game. And there’s no way that’s going to happen, especially at this pace.

Whatever game I make tomorrow, it needs to have only one core difficult thing to implement that the gameplay will revolve around, and the content should neatly wrap around that gameplay mechanic without needing to create a bunch of assets or extra code to give that mechanic meaning. If I can manage such an idea I think there’s a reasonable chance of success still. Most of the boring code should be out of the way now after all. For the rest of the night I think I’ll just add bits and pieces that might come in handy or try and fix some of the dodgy things happening at the moment.

The theme is cool, but so far I haven’t had any ideas that lever it to its full potential and are self-contained. Something about what lurks beneath the surface of people would be interesting, but short of making a game with lots of text and junk I haven’t had any viable ideas in that direction.

Alright, I’m superlate in starting, but that’s the way it goes.

Dropping this code here and saying it’s basecode: http://www.mediafire.com/download/h61lj5y3n3kxf6s/src.zip

I want to cannibalise  the movement stuff there mostly.

I’ll be using Haxe for code and gimp for graphics.

Not sure if I actually have enough enthusiasm to make a decent go at this, but don’t really just want to go and spend the day playing videogames either … although now that I say that it does sound like a rather nice idea.

I’m in.

Going to go with:
C++/Cocos2d-x (probably a poor idea, but I need experience in it)
Gimp
Tiled, if there are tiles.

Guess that’s about it. Hopefully I don’t fail as completely as last time.

Challenges that might be faced:
I’m going to a party on the Saturday night, which I fear will be all-you-can-drink. I may have to moderate myself, but if I don’t … well, we’ll see.
Also, my new ISP is apparently going to call me around 1pm and babble to me in Japanese, which will probably be an awful experience. But hey, anything to save a few yen.

So it’s nearly 12am. GMT+9 is a nice timezone to start Ludum Dare but a pretty awful one to finish it in.

I’m feeling fairly committed at this point. I’ve been pondering giving up on aiming for the compo and giving myself the extra day with the jam – it’s not like I have anything better to do tomorrow – but it’s not really the same.

The party last night did not help at all. Getting to bed at 2:30 am after drinking a fair amount of sake (language failure – we ordered one bottle between us which turned into one bottle each…) lead to getting up late and being pretty drowsy all day. This wouldn’t be the first time that a drinking party has scuttled my Ludum Dare efforts. It’s not like I go drinking very often even – it just seems that these things fall at inconvenient times.

But anyway, about the game.

The weapon is a bouncy ball. The idea is that you’ll run (jetpack, hopefully) through a level and bounce it around walls to kill off the enemies.
Currently the ball can be fired by aiming with the mouse, it bounces around, and it kills the enemy if it hits them. The player can also move left and right.

The art is horrendous (you’re not actually meant to be playing as some creepy hovering doll) and there’s no level design as yet. Still plenty of features that need implementing too – the enemies need to be at least mildly threatening and the player needs a way to get over the walls.

Even if there’s not all that much there, I at least feel like I’ve done a lot. This is my second project using Cocos2dx and I’m using a lot of stuff that I haven’t touched before. At the very least this will be a good learning experience

It’s  about 6am here and everything still sucks.

Well, okay, it’s not all that bad. Most of the game-stopping  bugs have either been fixed or have had the design adjusted to accommodate them. The only truly critical things left are a game over and a win condition. There’s something wrong with the aiming but I probably don’t care enough to fix it. And the player sprite desperately needs colouring or something.  But other than that, there’s a game here.

I submitted something!

http://ludumdare.com/compo/ludum-dare-32/?action=preview&uid=7956

This is only the second time I’ve managed to submit something playable, and it’s better than the last effort. I have fairly low expectations for how it will rate because, well, it’s  not very good or very finished, but I’m pretty happy with how much I got done. Cocos2d-x is pretty nice to work with except when it’s not, but I only had one really big coding frustration.

Anyway – it’s 9am. I’m going to eat breakfast and go to bed.