Docs

Docs

I have a few projects. Here I talk about how they came about and what I learned as I did them. Not a particularly technical segment, so refer to the sidebar for specific project documentation.

My software engineering

I would not really consider myself a software engineer. It is true that me being able to create a software project that solves a non-trivial problem, such as a game of blackjack, and being able to compile the project, puts me above many people in terms of software engineering skill. People who overwhelmingly do not know what a compiler is, what a programming language is, or even how the binary numeric system works. At the same time, my relationship to software is largely instrumentative. I am interested in software so far as it is able to solve issues for me, not because I care about software engineering in itself.

At the same time, software is indistinguishable from magic, and its usefulness cannot be overstated. I cannot help but be somewhat passionate when I see it being abused for the most idiotic of purposes, such as Electron or the GNOME desktop environment. Just because I have 32 GB of RAM on my computer does not mean your software is now allowed to use all of it. I still expect the software on my computer to adhere to the UNIX philosophy. Keep it simple, and you shall solve many problems. Electron does absolutely nothing except bloat your bad program that is just as bad, if not worse, than programs before it.

In summary, I think you could say that I am a software engineer so far as I am a computer user. I have work to do, I have a problem, and I need the right tools to solve the problem. Software that does not adhere to the UNIX philosophy is generally just bad software. This is why I care about software.

nimjack and nimchain

These were my first ever software projects that I would consider complete projects. This was by no means my first encounter with programming, but I never did anything that could be considered in any way a minimum viable product. Granted, nimchain does not have the essential component of networking, but from a design perspective it would not be difficult to plug into what I created, especially with Nim.

nimjack and nimchain were inspired by Nim. What I mean by that is that Nim opened new ways for me to think about programming. Before, many, many years ago, I would start with PHP, then move onto other “easy” programming languages like Python. However, I never really got the languages. Nothing ever stuck to me. I could never look at a problem and know exactly how to solve it. It would always be a matter of trial and error. This is because I never fundamentally understood key concepts of programming that are required to truly understand and appreciate programming.

Nim is amazing, because it has all the niceties of modern programming languages, with a syntax that doesn’t make you feel like you are reading with a needle stuck in your eye. At the same time, it is a statically typed language, where you have to reason about what you are doing in terms of type. Once I understood types, everything started becoming clear for me. I now could actually start programming, rather than fumble my way through and hope that one of the things I try succeeds. Consequently, I wrote some thoughts about programming languages, if you are interested in seeing my perspective.

bibel

This one was my first big project in Rust. It is not really a particularly good program. It was in fact first written in Go, but then I went onto Rust, and went all-out on trying some difficult programming problems to solve. The biggest one of those was implementing parallelism in the program, although that was accomplished using the Rayon library. It makes it very easy to do loops and many other routines as multiple threads, but you still need to reason about when and how that can be useful. It was an excellent learning experience.

Programming in Rust also taught me even more about what goes on in a program at a low level. I think Rust is the real C++, even if people give it flack for having a massive toolchain and compiling times. I always found it to be a weird criticism, considering that you would use neither Rust nor C++ to make small programs. You use either of those to make huge, complicated programs, but only one of those is really suitable without making it an incredibly costly and risky endeavour. I know at least I learned a lot from the experience either way.

Last updated on