Welcome to Hardmo.de. Enjoy your stay!

About the Author I'm Andreas Molzer, fascinated with distributed systems and the mathematical foundation of machines. Outside of coding, you will find me tinkering with electricity and hardware and gardening. There's no better tomato than a home-grown one. I'm residing in Germany, which means a timezone CEST which is UTC+01 in winter and UTC+02 in summer. Reach me on Github, contact (at) hardmo.de, or shoot an e-mail to the address in one of my commits.
Paradigms and Open Source Most of my projects are related to backends where I focus on clearness, clarity and conciseness. Besides experiments in programming languages, contributions to some open source projects have been made while I used them myself. The biggest so far to a library used in Mozilla's Firefox Quantum: image. It fixed several bugs and provided methods for decoding all of the image formats bundled in libnetpbm. The libraries I develop recreationally are almost always licensed as open source, mostly with the MIT license. Over the years though I've gotten a little more cautious with releasing things. Generally, the libraries will stay private until a sufficiently good demonstration is available or the potential to positively affect the workflow of other developers is clear. Learning is Strength. Freedom is Liberation. Understanding is Peace.
Security with Rust Since the first time I tried it out, Rust has stuck with me and as I planned on revamping my webserver (to the version in front of your eyes right now) concepts for security in distributed applications have sparked my interest. After I had unsuccessfully tried to find a server side implementation of the OAuth2 protocolmy search included other languages I'm fluent in as wellI decided to simply write my own: oxide-auth. (O xygen is the oxide in the formation of rust. The fact that the project was initially intended only for the web framework iron makes this pun even worse.)
What is there to see? A definite show-off is the following project in C++, turning templates into a complete functional programming language. While this might seem mind-bogglingly useless at first glanceadmittedly, the end result is merely spending a lot of RAM and time on a problem that could be solved easily by a code generatorit also highlights two key aspects of modern type system designs: they are extensive enough for providing expressiveness beyond the builtin capabilities of the language; and a committed programmer can use them for additional, generated, compile-time interface contracts not forseen in the C++ initial design. The project has been explored in code as well and is available on github.