I occasionaly work on some programming projects on GitHub I find to be either useful or interesting. Unfortunately they usually end up in the latter category.
As part of my secondary school project I’ve developed a web application that compares a selected set of pathfinding algorithms against eachother in randomly generated mazes. For statistical purposes it measures execution time, iterations and path length of each algorithm for each run. The web application is accessible here with its accompanying thesis document and statistical analysis here (in german). The entirety of the source code is freely available under the terms of the GNU Affero GPL v3.
Ever seen those videos where sorting algorithms are visualized either in bar charts or similar graphical visualizations? I figured this concept would be neat to apply using image data. The result is a very crude CoffeeScript implementation of said idea. Only one sorting algorithms is implemented. That ofcourse being bubble sort.
The idea would ideally be reimplemented in a faster compiled language. The web platform does not suit itself too well to this kind of application, though it does make it easier for people to discover and demo it. Perhaps C with SDL2 would be better tools for the job.
I programmed some trivial fetch’n’forward Telegram chat bots some time ago.
A front end to the Searx meta search engine (GitHub Repository).