projects.

Hereʼs a collection of my personal programming projects, both new and old.

---

meme better

Thanks to the popularity of sites like Reddit, a meme image (or image macro) has become a popular way to spread a joke amongst friends. To make a meme, I searched for a site that could quickly generate memes from any image, and without having to fuss with uploading it to some random personʼs server. Since that kind of meme generator didnʼt seem to exist, I made one.

meme better is a client-side web app that runs in all modern browsers (including mobile). It leverages HTML5ʼs Canvas object to generate memes on-the-fly within the browser. Users can start from a pre-loaded set of image templates, or they can provide their own image without uploading anything. The user applies their own text and when theyʼre happy with the image, downloading is immediate.

---

backstage

I set out to create a digital “hang out” for a small community, similar to a social network, but with more a friendly, and exclusive feel to it. The idea was to progressively build on the set of features, enriching the overall user experience over time.

Of the now many features, the first feature developed was an AJAX-driven live chat for online users, and a notification system to allow offline users can catch up on member activity the next time they log in.

The most complex section is the Image Gallery and Meme Generator. Users can upload images, then subsequently crop, resize and caption them as needed. The work done in this section became the base from which I created memebetter.com.

I really challenged myself with this project, tackling front-end and back-end development, and (some) design. I put obsessive care into user experience and consistent interface so that the siteʼs features are easily understood, despite their technical complexity. Itʼs a fun side project which I add new features to any time Iʼm inspired.

---

Vault

All too often, web companies need to share a large number of usernames and passwords. All too commonly however, these password are stored insecurely in a shared document, or worse, everyone has their own storage method. Vault was a holiday-break prototype which turned into a full project. My solution to problems mentioned (with some design love from Plank).

Vault was a host-proof, in-browser encryption tool with server-side persistence. It allowed safe storage and retrieval of sensitive information (such as usernames, password, software licenses, etc..) between a private group of users.

Vault had been in daily-use at Plank since 2010, and managed over 600 shared assets and credentials.

With so many well-vetted password managers out there now, you don't need to make your own. This was a fun project to serve a need at a time where password managers were largely offline and hard to share with a team.

---

role-playing video game

This project was my first real venture into video-game development. I used a basic low-level library SDL, and built everything else on top. I ported what I retained from J2ME mobile game testing and bug fixing, and then taught myself the rest along the way.

I very much enjoyed creating the basic game engine, human interactions, and laying out the gameʼs graphics. When the need to generate a map/level came up, I designed a map editor to make things easier. This allowed me to stitch together multiple levels, choosing the enemy placement, type, and map exit points.

Iʼve always had a fondness for the Super Nintendo-style graphics, so the 1980ʼs gamers amongst you will recognize the Mystic Quest, and Legend of Zelda sprites in-use.

I sat on this code for a long time, until early 2015 when I decided that it should to be open-sourced. I gave the 7 year old code a clean up in order for it to compile in more recent IDEs under Windows & OSX before releasing it into the wild. Have a look at the GitHub repository for a more detailed account of the project.

---

3d tank game

I wrote this after having tackled the 2-dimensional game, so I figured, whatʼs another dimension? This is the code base which gave me my long-standing respect and admiration for professional video-game developers.

The basic premise of this “lite game” is to navigate a maze, trying to find an exit point. Once reached, you can engage in a shooting-gallery-style battle with other tanks.

The coding challenges that arise from a real-time, 3d game, were more complex than I had expected and required a full toolset of Computer Science concepts. I worked long, tireless hours tweaking hand-coded 3d cameras, collision detection, map generation, and 3d vector drawing.

To make things no simpler, the basic requirements for this project were: no 3rd party modeling tools, and no gaming libraries. This meant every object had to be drawn out programmatically with the coordinate system in 3d space. I definitely learned how valuable a time saver 3d modeling software would be.

Although this is not much of a “game”, I was very pleased with how it turned out, especially having been my first time working with 3D.