💾 Archived View for soviet.circumlunar.space › zwatotem › diff › mockups.gmi captured on 2024-12-17 at 10:31:25. Gemini links have been rewritten to link to archived content
⬅️ Previous capture (2024-05-10)
-=-=-=-=-=-=-
I should really be working on my master thesis right now, but I have a phylosophy going by the way, so I won't miss the opportunity to write it down.
It's really not that easy to write a 50 page paper. I am worried I will not be able to, but at the same time I am not worried enough. To be fair, I deliberately chose this topic. I thought it is a real computer science, reaserch and development worthy topic, more useful to the society than all the copy-cat works of my colleagues, that will be repeated by someone else just by chance, or is unique, but does not bring significant novelty. But even if this is true, not single of my dependencies... references is longer than 20 pages I think.
But anyway. That all was to say, that academic work is hard, or at least non-trivial. I've been made aware of an article recently that criticises the composition of universities, and their goals. There is a great rush in publishing articles, even if they are of no notable quality, contrary to past times, when scholars had all the time they needed to properly develop their work.
The work I read recently was on a system that transforms code for sequential tree implementations into parallel ones. The algorithm of transformation was laid out in plain english, but no transformer was implemented. It was "in the works", but the GitHub repo was a 404, so I don't know what's up with that.
Popular phrase "Further work is needed" sounds like the author didn't end what he started, for a lack of time, for the pressure of publishing right here and right now to get a grant, or get on this year's conference. Could even be for lazyness, and lack of passion. Researcher not by vocation, but in spite of themselves, delivering a minimum viable product. This seems sad.
In on the topic of my work though. I read this good paper, and it mentions some other papers, where the authors created very interesting ways of writing programs, but their build targets was... Java.
At first I wanted to criticised that, because I knew that this is a reason that we can't have nice things. Every unique idea for a programming workflow (I deliberately avoid the word language) that comes out of academia gets implemented in this half-baked way, where it's a weird transpilation to Java, or some other language. Or it's an extension that adds a lightweight syntax to a language. This is good for a first prototype, but programmers are lazy. Having one more build step is one too much. There is enough jank in programming process already. And unfortunately if the original author doesn't do it right, noone really wants to be that guy, and take on themselve a hard task of implementing a whole new paradigm of programming in an end-to-end stack.
But then I remebered, that I'm going in the same direction. Cutting corners, just to make it in time to submit my thesis. Instead of custom language, or even a propper one with strong type system, rendered in a nice UI, actually stored under the hood as an axis-agnostic data structure, compiled to LLVM let's say, I'm using Julia, completely dynamically typed language, trying to manually restrict it, then add some extension syntax, then take a subset of it, because I won't handle parametric polymorphism, display it in a grid, sure, but absolutely store it as code, and then probably run it in an interpreter, just like some lame Python. Pyhon is not even in a set of languages that are affected by expression problem (it really depends on the definition, but let's not dig). And am I even going to perform that "future work"? No, of course not.
After I get master, I immediately become a corpo-rat, and forget about fundamentals of programming. And this paper is not getting read by someone and picked up on either, I can promise you that. Even if I pasted it on my page
Oh, there it is, my page (mostly down for now)
Even if I did that, that would technically be illegal, as I'm pretty sure my University contractually takes the rights to publish my master thesis for 2 (?) years after I submit it. But regardless, even if I did it, good luck having a targeted reader on that that will do something about it. It's a miracle already, that I stumbled upon "Programming without text" article and that inspired me to do something about it. What are the chances that this effect chains?
Programming without text (archived version, because the original is down)
So that probably excuses why I am not particularly freaked about my master thesis. I think, that I should either do it well, or not bother. In the end it all is just a paper. I could as well write it down if I had the time. Does not doing it make me any less of a CompSci master that I am? Does it nullify all the work that I did and all the things that I learned here? No. But unfortunately to the world a crappy paper with a master stamp on it is worth more than an excellent one without it.
But if I don't make it, my mother will disintegrate, and I can't let that happen...