4 Years of game development


Nov/06/2011 was the day I released my first game to the public. It was named Kill All The Marcos and was played just by a few people. It was a game inspired by the meme X all the Y from Hyperbole and a half.


It was a bullet hell game in which you shot and kill spawning enemies before they killed you. It was created in game maker 8.1 and it felt like an expensive purchase at the time. I didn’t knew I was buying my ticket for game development.


I decided to start with a bigger game. I had many ideas of games I would like to play (and make), the best ones:

  • A zombie survival game in which you protected a wall from above. You could decide to help a NPC at the cost of your food and gain their help or just let them perish.
  • A game about a being a Dragon.
  • A mobile game in which you controlled a mage and draw symbols on your screen to cast spells to protect the castle.


The game I started working on was called Holgoryon. That was the name of a magic book which contained the instructions to build everything. The world makers would rip pages from the book and mix them together in a cauldron to build a new world until one day they dropped by “accident” the whole book in the cauldron.

I tried to recruit several friends to help with the game, but failed. Started working slowly in the game with some concepts and illustrations:



This was my drawing stage. I already knew how to code, but graphics were a problem. Holgoryon was a very cool idea, but too ambitious and I had to drop it. This failure helped me notice I had to work in smaller projects and it is when I discovered Ludum Dare.


I started with LD23 with a game called A Bathtub Story still using Game Maker. A bullet hell game where you played as the bathtub monster, destroying ships to clear the level.


It felt great to make a game in under 48 hours.


This year I tried to get new proyects running and continued drawing.


The concept of Helldust was born:

helldustlogo-1.pngIt was supposed to make us better. We were promised a ticket to heaven, but we were just purchasing our doom.
At first, we had more energy and our brain power increased. Soon we didn’t have to sleep anymore.
We bought more, we used more.
The compound was known as PGB-75.
We now call it helldust.

It was an RPG game based on using chemical enhancements instead of magic.

helldust2.pngIt remains as a concept, but development never really started.


Skipped some Ludum Dares, but returned for LD28 with Monster Interceptor with a pulsar ray that allowed your ship to travel really fast. You only had one chance to save humanity. Still made with GameMaker, but discovered how to export it also to Mac.

11246-shot0.pngI was really happy with this game and tried to make it a full game. I called this remake The Erebos


This year was huge for game development and there were so many proyects.


Decided to get some professional help and hired an artist to continue with The Erebos now called Konrad. I really enjoyed working with a professional illustrator and we had some cool artwork.


Later I had some doubts about the proyect. I felt the game was not fun enough and cancelled it.


Another participation now for LD29 with Switcher. It was really well received and was played more than 1000 times. It is runner game where you are chased by superheroes as you are a monster and are considered a threat.



Using #lowrezjam as a motivation I coded Gozira in a 32×32 pixels canvas. It was a cool restriction and really enjoyed making this game. The goal is to avoid getting hit and get the highest score you can.



Another game jam, now #spacecowboyjam and I made Sand Rangers a 2-player coop game in which you capture bad guys. A bullet hell game with beat your high score mechanic.



Another LD game, Mammoth Monkey Mole. A puzzle game. This time I got help from a friend and we released a really complex game. Control 3 animals, each with different abilities living in a different dimension.



This time I was hooked by a challenge. Make a game in less than 13 kilobytes! The js13kgames challenge. I delivered Shield Maiden. A rogue-like game to capture the 4 horses of the apocalypse to prevent it from happening. I learned a lot from the size restriction.



The original game felt really good and the iPad version started development.



The last game of 2014, Paws of Fury. A game in which you fight other cats for the throne. Just like Switcher it was played more than 1000 times, a really nice surprise.



Duck Attack

Inspired by a meme (Karl King of Ducks) made a game in which you control ducks by throwing bread to point where they should attack. The best thing about this game, is that someone like the game so much they donated $1 my first dollar as game development. Play it here.


Eat Sheep

I really like puzzle games so for Ludum Dare 33 I made a simple game. You control a monster and your goal is to eat all the sheeps before they can hide. Play it here.


Fantastic Monster Chronicles

My second time participating in JS13k games. This time I even had to remove stuff to fit it into the 13 kb requirements. I even found a tool to generate music and I composed it by myself. Visit the js13k games site to play (Use chrome).



Mammoth Monkey Mole

This year was almost completely dedicated to finish this game for iPad/iPhone. The game is entirely coded in Swift and I am happy with the end result. The only drawback was that I was completely exhausted by the end of the year, but did learned a lot.



Download free from the App Store.


The year is just starting and I have new ideas. Haven’t yet decided on what to commit. I have new ideas on more puzzle games and I am hoping to try new concepts and new graphic styles.

The designer developer

As a full-time iOS developer, most of my time is dedicated to front-end, dealing with designs and implementing them in an App. Someone designs and developers make it real. It works most of the time. We could say both dev and designer need to speak “UI” in order to interact between each other.

One way to improve communication is to have a designer learn about programming and viceversa. I tried with the classic designer tools: Illustrator and Photoshop. BUT they are way expensive! A free alternative is GIMP and Inkscape. GIMP is ok, but Inkscape it kind of clumsy.

Finally I learned about Sketch. Is a design app focused on UI. It feels intuitive and at the time Sketch 2 was only $49 dollars which is a fair price for all the available functionality. I designed some UI and started pushing how far I could use the app.

I was watching an anime, Last Exile, and decided to do a quick vectorization of a character. It was quite easy and the results were better than expected.

Tatiana TatianaBase

As a result I started making my game art in Sketch. Mammoth Monkey Mole iPad remake is all made with this tool and I am happy with the results.

December Ludum Dare

48 hours to make a game. An awesome idea.

I make games for two reasons:

  1. I like programming them.
  2. I like that people play and like my games

For reason #1, is really simple, you sit, program and done. But for the second one, getting people to actually play your game… that is quite difficult. It is even more difficult to get people like your game.

Ludum Dare is perfect for #2 reason. It has a huge audience and if played right, you can get a decent amount of plays in your game. With people playing your game, you get feedback and feedback makes you and your game better.

I started with Ludum Dare 23 using game maker to build a bullet hell(Bathtub Story) and then took a huge break and made a comeback in LD28 with still more game maker and made Monster Interceptor. Then LD29 with javascript programmed Switcher which to my perspective was a huge hit and even got a mention in Indie Statik. And the last one LD30 with Mammoth Monkey Mole (a modest hit) also in javascript.

Every game made in LD has taught me something and that is priceless. Looking forward to LD31.

Bug fixes

Before new features were implemented it was time to solve all the existing bugs and avoid future trouble.

The issues in iOS 7 with Swift were finally solved. A SKNode layer was troublesome by capturing the touch events and the z-ordering was not respected. The downcast issue was also solved by avoiding the cast and work directly with SKNodes instead of SKSpriteNode. They were really a pain to debug, but with those solved I could focus in the gameplay.

Levers and bridges were implemented, as well as the valid movements for the monkey.

iOS Simulator Screen Shot Dec 2, 2014, 9.19.51 PM

MMM Updates

Implemented the functionality to pick and drop the Dirt-balls. Also both rocks and dirt-balls can be moved to cover a water hole, so they have their own sprite to show what has blocked it.

The bad news is that iOS 7 is not handling correctly  the touch events. The swift downcasting from SKNode to SKSprite node is failing and thus sometimes the dirt-balls are not being able to be dropped.

Other issue is that it may not be completely clear which dirt-balls can be lifted and where they can be dropped. Spend some time designing the icons to avoid that issue:

mmm_2nov_2 mmm_2nov

MMM Game progress

The MMM game was originally created in Obj-C and SpriteKit and was generated using a tutorial for a candy crush clone. It made sense at the time because it was tile based and MMM is tile based. I was wrong.

The code base was a mess and only noticed it when tried to extend the functionality and it was a complete pain.

Lesson learned: Tutorials are great, but design your own architecture for your specific needs.

Not all was lost, since I get a better understanding on SpriteKit, and some code could be reused.

After finally unplugged the old code and left it to die. I decided to take my chances at Swift instead of Obj-C. I am using SpriteKit, which is only supported in iOS 7+. Changing to Swift gives me the opportunity to update to new technologies and also embrace the change.

One week later I finally feel like I am been productive again. So much type casting is still a mess in Swift, but you learn to live with it.

The current version can:

  • Collide between animals or board objects
  • Change between active animals by tapping on them
  • Move using swipes
  • Mammoth can move rocks
  • Holes act as barriers

Here is the current build with a screenshot from the simulator:


It runs weirdly slow in the simulator, but in a ipad mini retina it can run at 60fps without trouble.