During the Dare weekend, I made a little adventure game engine.

In the course of doing so, I also went about making a little adventure game about a girl that just wants some peace and quiet.

Play the Web version.

Ludum Dare entry page.

We are almost there !! We are now including all the images to the game’s engine.

A demo ? Yes, here.

Game’s Engine with images.

We are about to be there !!

Here’s my time-lapse of making Spawn this weekend. Total capture time was about 21 hours. Condensed to 9 minutes.

I forgot to put up my big clock for the first hour or so. I started recording at 6:45.

View the Spawn Timelapse

Play it here

This was my first ludum dare entry. I made a underwater surival game. Here is the mini post mortem:

What went right

Framework: I made some games with libGDX before, so this was a good choice for me. I used the experimental model loader extension but besides being experimental everything worked as expected.

Models: I used Blender for the modeling and texturing. All textures are created and baked in Blender. Because of my limited modelling abilities and time constraints the models are kinda low poly but this fits quite well to the overall art style. For the animations of the shark and the seaweed I used a simple wobble vertex shader. This was a huge time saver!

Sound: I recorded the sound effects with my phone or BFXR and added some slomo effects and additional bass with Audacity. This worked quite well. For the music I used some generated tunes of autotracker bottomup and exported them to midi. I imported the midi to LMMS and used a funky synth. Some additional slomo effects in Audacity resulted in the final ambient music.

Mood: FPS + Fog + Sharks +  dark ambient sound = WIN!

What went wrong

Game play: The overall game play is to simple. I got no real idea, so I just started a FPS with no real game play target. The initial primary goal was to collect stuff in a limited time constraint. This was to simple, so I added sharks to the game.

Collision detection: Spend way to much time on this one. I did something wrong with the model export. The collision center of all models had a wrong offset… The final collision detection is still kinda hacky in the final entry… :/

Overall I’m happy with the result. Good mood in the game and I learnt something along the way. See you next time!

I just found out my LD#22 submission isn’t showing up in the list or is even searchable… if anyone running Win7 wants to play a half-baked demo with no win condition, here’s the entry’s direct URL:

AL One:

http://www.ludumdare.com/compo/ludum-dare-22/?action=preview&uid=8105

My submission is no longer bound to a zip download!

You can now click directly on the ‘web’ version, and actually get a web version! wooo!

Alright, I managed to beat GM4Mac into submission and actually port my game, so now Comfort Zone can be played by my fellow Mac users.  Hurrah!  This means all of my Ludum Dare work is done.  Now, to play as many of the other 774 entries as possible!

Comfort Zone Postmortem:

The bad:

I spent most of my first night messing with movement variables, trying to bring the central gimmick of my game to life in a way that both felt “true to the idea” and fun.  I ended up with an overcomplicated mess of code that didn’t communicate anything when you played it.  Then I overslept into Saturday, and woke up feeling disheartened.  I wasted
most of the day searching for renewed inspiration.

The good:

During that search, I created a short song that really helped nail down the feel of the game for me.  I simplified the movement code and the gimmick, so that the idea was communicated consisely and powerfully. During the last 12 hours (I didn’t sleep at all during night 2 )  the game design finally started to click. When I was done, my head was a mess and the game felt like a jumble, but I’m pretty certain I managed to piece everything together.  It’s short, it’s weird, but there’s something fun in there, and something interesting in there — and I consider that a success.

Sound effects:

Speaking of clicking, I did sound effects differently this time.  Originally I planned to use Bfxr (Stephen “increpare” Lavelle’s upgraded version of Sfxr), in standard LD fashion.  But then, when planning what sort of sounds were needed to communicate, I had the idea to use my laptop’s built-in microphone, instead.  Like I did when I first created Game Maker games as a kid, I would record a short, low-quality clip of something hitting something else, or scraping, or my voice.  Then I would use Sound Recorder’s basic tools to modify the clips into what I needed.  I think the results were perfect, especially for this game!

Theme:

The theme is the most important part of Ludum Dare, to me. The time-management challenge is fun, but the game-design challenge of an unusual theme is fascinating to me.  I always try to have my game mechanics inspired BY the theme, instead of creating a game and then “working the theme into it.”  Ironically, I think this can result in a game where the theme is less obvious.  This time, my thought process went thusly (minor spoilers): “Alone –> your character is alone –> aren’t game characters usually alone? –> your character ISN’T alone, but wants to be –> gameplay involves avoiding people so you can accomplish tasks”.

Graphics:

Well, I never found time to change my placeholder art into something more elaborate.  So instead, I just ran with it, and tried to turn the style into something communicative.  I think it worked out alright in the end.  It’s usually pretty effective to use base colors to represent different themes.

Music:

Probably the most “polished” part of my game, you can download the 53-second song here if you’d like: http://www.host-a.net/u/Ditocoaf/alone.mp3

Thanks for reading!

I’ve finished my game, and now I’m wondering where would be a good place to host it (I’ve never bought web hosting). Its made of HTML5 + Javascript. It’s 7mb. I’d be nice if it was relatively fast. I’d be willing to drop $10-$20 for a temporary hosting soloution.

Playable here

Last time around I had jumped feet first into flash and actionscript3, knowing very little, but learning a lot. This time I was going for a way more graphical project, and I was going to make it within unity.
The theme ‘alone’ hit and it meant I would be doing a space themed game, with no other people, and a lone astronaut. I love that kind of sci-fi.

The good

Using unity.
The Unity engine turned out to be the most useful thing this time around, the amount of prototyping that could be done rapidly, the seamless exportation from blender which meant I could make a change to a model and it would instantly appear back in the game.

Graphical tricks
Most visual things in this game is pulled off with sloppy tricks, like focusing the detail I made where you actually see it on the player and leaving other areas largely untouched. Textures were made less to ideal, but still functional, so that I could get them in and pull a consistant visual rather than just ending up with two models.
All environmental station parts are built from a tiny piece of texture with enough variation to pull it off, the outside is nearly a crime, but works. Leaving the shading very flat overall gave me the ability to paint in more details into the textures themselves rather than having to rely on shading, which would add more work on top of the much I already had on my list.

Modular reusable gameplay elements
I made a few objects and wrote scripts for them in a way that I could just spread them out and connect them to things, such as doors to any amount of hacking panels, which at the point of writing the story I was able to squeeze out a few datapads within the last hour or so, and alter around hacking panels for different effects such as hotspot movement/shrinkage, difficulty, and amount of hacking stages.

The terrifyingly bad

Late actual prototyping
The first day was spent entirely just thinking of a setting and making art, with the first thing being a fully textured player model, could’ve only been worst if the first thing I made was a menu. The effects of this can be seen in the game itself, while the setting worked out nice and looks pretty, the gameplay is rushed and largely untested, with the hack-minigames lacking the variety they could’ve used. Prototyping early will be my goal for the next ludum dare.

Heavily visual project
Aiming to make this very visual and 3d project was a bad idea and still is, but still totally worth it. The effects were that of spending a majority of the time on art, which essentially makes the game an interactive story rather than a game.
Making a simple 2d project will be my goal for next time, but I know I will most likely end up doing another full on 3d project.

Lack of testing
Large parts of the game and story was implemented late the last day, which means they went untested, I wasn’t able to give it a proper gradual increase of difficulty, and instead it jumps around.

Bonus

Overall I was quite satisfied, from learning, pressing myself and eventually not giving up, which I was close to doing early the second day. My marriage also survived another ludum dare!
Definitely doing this again next one!

So originally I said “I will not use a object-relational mapper.”  My reasons were:

• Overhead
• I don’t mind designing by hand, in fact, I’m fairly good at it
• I don’t have experience with any aside from django’s mapper
• I am not using django (waay too much overhead and wrong toolkit for this.)

I’d looked at the Canonical thing, I think it’s called storm, and it looked decent but in my attempted use of it a month or so ago, I hit upon lack of documentation beyond the tutorial which did not explain some basic things, like “how to get your db design to map.”  Sort of, you know, what a mapper… should do.

I do not feel that sqlalchemy is a good choice because it’s just plain too much overhead.  I don’t really need an ORM.  I need connectivity, and I’ll handle keeping the database in line. I’ve dealt far too many times with software that assumes that if you use a mapper, the database will never get corrupted. Like, you know, 100% of all java db stuff I’ve ever dealt with. In production. Costing me months of time. I hate java; too many “developers” who _can’t._  Those who _can_ develop in java never end up working anywhere near me.

Anyway so with all that background, I said “no” and was busy setting up the db by hand, and – darn it, I went and changed tools. I used to draw stuff up and build the whole design first, then publish; but no I went and was doing it bit by bit using Sequel Pro. No. Bad. It doesn’t … it works but it is NOT a design tool. It is … usable for some stuff I’ve been dealing with at work. With crap db’s designed by people who have no clue. _should have known_ but didn’t think it would be THIS bad.

MySql Workbench worked well but lately it’s just been PANTS. Meaning “Hi, I will crash now!” and such. Also they don’t do ppc. And… yeah.

So in short: I will need to probably write my own ORM type thing after this is over. What I need is a way to define the db directly, quickly, and without having breakage upon a change. And very little overhead.  I want data, put into an object, and I’ll do the push/pull when I need to.

Oh that’s right, another issue with storm was it did not explain in any way, shape, or form how to avoid the problem of persistent connections that needed to be re-attached. Like, if you went to sleep, and they timed out, how to reconnect them the next morning when you were looking at it.

Anyway not publishing this til after. Just saying, this is one of the things that went awry.

Ok seriously, I need synergy working with osX ppc and Arch Linux intel. It doesn’t. This swapping the usb keyboard back and forth is SERIOUSLY annoying, and since I suck at maintaining concentration, it REALLY gets bad.

later

I realized, suddenly, “Hey I tried brew instead of macports, I wonder if it has synergy and if it would work, I mean… this swapping USB keyboards every 3 minutes is REALLY bad”

It did.

IT WORKS

I AM SO HAPPY

I just submitted a jam version of my game!
I added sound effects, menu screen and some minor tweaks.

PLAY IT HERE!
LD entry page

PLAY SPAWN

I didn’t keep very good notes while I was developing this weekend, so I’m going to try to just list out approximately what happened when before I forget too much. I’m referring to my time-lapse video to help me with the exact times.

Friday

4:00 pm – Took a short nap and then bought some groceries so I would be well rested and fed for the upcoming competition.

6:00 pm – Theme is announced! ‘Alone’.
Initially I am not too crazy about the theme. It seems a too broad. I imagine everyone just making standard platformers and dungeon crawlers and saying ‘the main character is alone’. I suppose that would happen with any theme.

6:00 – 6:45 pm – Brainstorming and sketching ideas.
Some idea fragments:

• the last leaf clinging to a tree at the start of winter
• the last chick left all alone in the nest
• a lone amoeba swimming through a vast ocean
• piloting a ship through empty space
• a dandelion spore

6:45 pm - Open Flash and TextMate and start prototyping the basic movement and gameplay

7:45 pm - Briefly consider using Flixel. Decide not to due to my unfamiliarity with it.

10:00 pm – With a rough prototype complete, I stop coding and write a quick blog post.
I spend about 30–45 minutes sketching character designs in my sketchbook.

11:00 pm Sleep!
Saturday

5:30 am - I woke up on Saturday feeling refreshed and was glad that my game concept and prototype still seemed viable after sleeping on it.

6:30 am After breakfast I immediately started working on the designs for my main amoeba character (in Illustrator).
After about an hour I had a basic design that worked so I brought it into Flash and started animating.

8:30 am - Design title screen

9:00 am - Change Nutrients from being Flash Sprites to bitmap objects that get blitted onto the background, hoping to improve performance.

10:00 am - Added the collected nutrients to the amoeba’s belly.
The game becomes pretty playable at this point. From here on I think I wasted a lot of time just playing my game.

10:30 am - Sound effects (sfxr & Sound Studio)

11:00 am - Added spawning animation.

1:30 pm - Added ‘New Species’ label

2:30 pm - Added some placeholder new creature graphics

3:00 pm – Added the menu for showing all the creatures you’ve spawned.

7:00 pm – Added background bubble graphics

8:00 pm - Remade title screen as multiple graphics instead of just one.

I was starting to feel really burned out at this point, so I wrote a quick blog post and thought I would go to bed. Instead, I spent another hour or so working on some background music (GarageBand).

Sunday

6:30 am - On Sunday I woke up feeling terrible. I was tired and my whole body was aching. I was not looking forward to spending another day in front of the computer, but I knew I only had a few more (big) things to finish.

7:00 am – The major thing remaining was to draw the graphics for all of my spawn creatures, so I got started on that first thing.
It took most of the morning.

10:45 am – Added different probabilities for which nutrients show up based on where you are in the ocean.
I took a few big missteps before I got this right.

12:00 pm - Fearing poor performance on older computers, I decided to change the game’s framerate from 60 to 30. Most of the movement in the game is framerate independent, but my amoeba animation had to be adjusted.

1:00 pm - Posted my finished game with 5 hours to spare! Woo hoo!

my head is on fire

there are some missing features and some bugs since we were running out of time.

http://www.ludumdare.com/compo/ludum-dare-22/?action=preview&uid=6685

I know you all are rating and sharing games right now, but today I received what is certain to be my best XMASS gift of the year:

See pics?

Suspicious postage…

Suspicious packages…

Supremely executed ribbon curling technique…

Open… fiiirrrssst??

The gifts are great! The cats particularly like the änglaspel

Thanks Swedish Santa and God Jul, everyone!

My games done! Well, yesterday it was. I was too goddam tired to make a post about it right then! Now after having a little rest, heres some kind of postmortem.

Heres the game itself: ASSAULT THE DARK FORTRESS

The game is a somewhat long zelda-like game where you explore a big maze, find upgrades, keys, treasure and food. Food goes down slowly as you play, and you can starve to death. You soon get a sword and other equipment to help in combat and exploration.
The game takes maybe 30 minutes to one hour to complete.

My biggest mistake about the game was probably how large the map was. I spent more than a day working on the games world layout, before even having visuals other than placeholders. I was finally done with the map when I had only 5 hours left on the deadline… at which point I spent one hour making a proper tileset:

Then I spent 3 hours placing these tiles in the game, and one last hour putting in sounds and correcting bugs. The 3 hours of tile laying wasnt enough for the size of the game however, I only had time to place tiles on the left half of the world map, so the entire right hand side of the game ended up with just a background color and dark gray blocks for walls! eep!

Another problem was that I had very little sleep on the first day. I went to sleep for 6 hours and only slept one of them, as a result, the second day was very sluggish, and I did very little compared to the first night.

All in all I’m rather happy with the result though. It has some bugs, like the dialogues of the intro sequence reappear randomly as you play (its a really stupid bug causing that, too!) And also a couple room exits don’t link up properly. But other than that it seems to work fine, I was able to explore the whole maze. Also the rush at the end forced me to somewhat do the tiling quickly, so many rooms look a little similar, which is a problem with exploration. Still, killing stuff is quite satisfying, and upgrading your weapon is nice.
I will surely spend quite a bit of hours polishing and finishing it up in the coming weeks, so expect to see a much better version eventually, with more varied art, less placeholders, better sound, more varied enemies. The current lack of music is disappointing, too, ill have to look into that.

Well this has been my first ludum dare, and I have to say it’s been a cool but exhaustive experience.This time I’m going to write a small postmortem for my entry, Enola, which you can check out here.

What went right:

Having a clear concept for the game from the begining.  The first thing I did was write the entire (tech) document on paper. I wrote what the entire game would be about, tech aspects and limitations. This prevented me from getting lost during development.

Using a tool I already knew. I’ve been using UDK for some time, and that knowledge helped me to better decide the kind of game to develop (in conjunction with the previous item), so I didn’t find myself trying to figure out how to do anything.

Simple graphics style. The theme “alone” can mean many things to many people, but many entries agree that “mood is important.” For Enola I used simple models, only a couple of textures, and some post effects to sell the look and feel of the game (like the thick fog, the exaggerated DOF, and soft bloom).

A simple control scheme. I wanted Enola to be the kind of game that takes as much as you want to finish. There is no combat or action sequences involved (so you can’t really die in the game), and that makes the game very inclusive so (almost) anyone who can use a keyboard and mouse can play it, allowing players to play at their own pace.

Using the environment to guide the player. When you start the game the first thing you see is a glowing red orb and when you grab it, it says something like “the trees are pointing in the same direction.” This was a simple way to guide the players, so even if they get lost, they can follow the pattern of the trees to find the way. On the next level, you get a “clue” from a painting so you know it’s important, and I think these clues serve their purpose well.

What went wrong:

Not being able to work full time. Yesterday I realized I’d only worked around 23 hours on the game. That’s because I had to go out a few times, so I couldn’t work as much as I wanted. Besides I got sick on saturday, and the pills I took made me drowsy.

The game is too short. Since I couldn’t work as much as I wanted, the game is too short. The second level was planned to be twice as big. I wanted a bigger house with more to explore and do. Besides, the bigger house would have added to the “mystery effect” where the house looks very small on the outside, but way bigger on the inside (like the Diablo chapel).

Not entirely happy with the audio. I really had to rush the audio “design,” and while the audio fits the mood, and the “noise” gets higher as you reach the goal, the entire set of sound effects could improve, a lot. All sound effects were recorded by me, using my microphone.

The game is way too short. I can’t stress that enough.

No time to optimize. The game is around 100mb, and that’s after I manually removed some of the things I was sure weren’t needed (for example, why UDK includes the Mac app builder when building the installer is beyond me). It also includes some other things like Shader Model 5 materials the game shouldn’t use since I was using DX9 exclusively (or at least I suppose they aren’t needed). Pretty much finding out what the game didn’t need so the download wasn’t 100mb in size would have been a good thing.

Conclussion:

If you ask me, I’ll definitely be in the next ludum dare, but I have to find ways to ignore real life during that weekend so I can post a better game overall next time. I should also find a better way to make sound effects.

I’ve also been wondering if/how the game could be made into a full game, depending on people’s reaction, since the concept can be extended very easily, and the clues you get from the game hint a larger story.

My submission to my first Ludum Dare is complete. I throughly enjoyed participating in the Ludum Dare.

The Submission page for the game is:

http://www.ludumdare.com/compo/ludum-dare-22/?action=preview&uid=6703

Enjoy!

What are those numbers in front of the names on the rate page?