in which i try to learn how to code

One of the things that bothered me a LOT about trying to learn tech is that a lot of the resources out there are geared towards getting a job. Like, I get it. But I'm not looking for a job, I just wanna learn cool shit and understand computers better.

Even stuff like CS50 is gonna be geared towards teaching you how to actually code, first, and that's a problem because if I don't understand how a machine actually runs code, then my brain is simply NOT gonna learn it.

I took a giant step back, then, and tried to find more info on computers in general, so I could actually understand how tech actually works, and would you look at that, I found the perfect way to do it: learning shit historically.

The crash course playlist is the actual place to start if you truly want to understand how shit works for real, instead of just how to make Hello World show up on screen. Finally someone found a way to explain to me how things went from basic logic gates to RAM, and CPUs, and made it make sense. No bullshit, no unnecessary complexities. Just. Here's how computers work.

It now makes a LOT more sense how things actually work, and WHY they work. The idea of levels of abstraction helps with the nesting of ideas, and it's life saver so you can wrap your head around things. In the end everything is just zeros and ones, but a block of zeros and ones that have a specific function is called X. A specific group of logic gates that do Y gets the name of Z.

For real. If you think computers are neat, wanna learn more but coding is way too high levels for you, this is the playlist you want.

In my books everyone should learn how things developed, and what actually is a CPU, and THEN learn to code.

Complete understanding or bust!!!!!!!! LOL

Still watching the rest of the first class, and found it interesting how they present the idea of writing the pseudocode already kinda thinking about the code itself, which is something interesting for me, that actually helps a lot. Less because I can actually do it, but because it can be useful to signal to yourself what to do if you don't know how to implement it.

The idea is that, in here —

— the verbs sort of work as functions, else/if are conditionals, "go back" means a loop.

It's simple, but in my mind I can already see this helping so much. I kinda need this more or less 1:1 relationship between a language I can understand and a language I'm learning (considering code like just another language like English, French, etc). Mapping those relationships when then exist is helpful to understand the logic of everything, and makes it clearer and more easy to understand when the relationship is NOT perfect equivalent or when there's no equivalent at all.

This is gonna help a lot actually.

And gotta love how they unmasked AI right out of bat.

It really is just complex conditionals, isn't it? There's a lot of impressive tech around it, but I have so much beef with AI. The ethics of where and how to use it and train it are still sketchy, and that clip of OpenAI's CTO purposefully failing to answer questions about Sora doesn't help any of my concerns.

Still, that cat is out of the bag, and the tech itself is useful and amazing, when used for good. It's a great assistant for humans, and can be used for learning in interesting ways, if you use it knowing that it's just a very good statistical machine — and that the answer that come out of it might not be true at all.

It makes me immensely happy that they made the duck debugger (lol) available all around. It's the kinda use of AI that I can get behind.

Finally, I'm also a little in love with Scratch. It's exactly the kind of "for kids" thing that I absolutely need to start out and feel like I can actually get this done.

This kind of thing makes me miss college a lot. I was kind of made for learning, and for academia, even if academia wasn't made for me, and taking this class has made me feel like myself for the first time in a very long time.

Not only I decided to continue learning how to code, I decided to take a more complete look into programing, and, really, computers and software. It took me to CS50, Harvard's beginner CS class that truly starts from the bottom and builds it up concept by concept.

I'm impressed by a lot of it, but first and foremost by how not disappointed I am. Which is, somehow, kinda sad. As a kid I had this dream of being a Harvard student, which faded for many reasons over the years, and but that still has a tiny hold on me. I like learning, I like learning well, and the idea of being in a place where people like it just as much was always appealing to me. Also, the aesthetics. The aesthetics was a big part of it. The years passed and it didn't happen, and I was glad in many ways that it didn't. First of all, because of money. I don't know about you, but I don't have that kinda budget. Second because institutions like Harvard have their own problems, and I was happy to not be a part of them. It's the part that disappoints me: that an institution like that is not well suited for the way I learn, and too reliant on the things that broke me as a student when I was younger.

It was a bit of a consolation prize, you can say. I never got there but, hey, maybe it's better like that. It was. It absolutely was. But taking CS50 from edX reminds me of why I wanted it in the first place. The class is, as the professor tells us in the first 1min, like drinking water from a fire hose. It's fast, dense and takes no prisoners, but it's also incredibly fun. Things are well explained; the logic of how the subjects are presented is flawless. It's human, and it assumes you're ignorant about the subject, but not an idiot.

I am, though, admittedly, insanely glad I'm not completely ignorant to the subject. I don't have formal instruction on computer science, and all my knowledge of it is patchy and inconsistent, but it's familiar, which does wonders in keeping me grounded on what's happening in the class. The first lecture is a whooping 2h long: 2h of non-stop new information. My attention span is laughable fraction of that. It is

Again: I'm glad I'm not actually at Harvard taking this.

I haven't finished the first lecture completely, I'm roughly one hour into it, when the professor starts introducing what I'm pretty sure it's gonna end up being big O notation.

I have no idea of how it works, and just a vague one about it is, and I was fading already, so I thought best to leave it for tomorrow.

For now, I'm gonna just stop and let my brain digest everything, and try to figure out the answer for a question that came to mind when he was explaining binary numbers that's bothering the shit out of me:

Why the fuck a byte is 8 bits, and not some other numbers? So far the internet has given me things that have to do with Unicode, and "it's just convention, which — fair, I guess. But one answer made me curious. Some guy on Reddit talked about the fact that the fact that 8 is 2 squared made me binary math easier, and memory better. I'm not sure what that means, but I want to figure out. Also, something in me tells me it might have to do with hardware... Back when all of this was being made hardware was a serious limitation, so maybe it's got something to do with it.

I always wanted to learn how to code, and the idea of code always seemed both very easy and very hard at the same time. Now, trying my hand at it for the first time, I realize that I was 100% right in my assessment. This shit is based on very simple concepts, but my mind was not made for math of any kind. Not even the very simple kind.

The way Khan Academy decides to teach people code is incredibly good, as far as I'm concerned. All the times I tried to learn code before it was always based on just writing things and hoping for the best and my head cannot deal with that. I have to actually see things, see results, see the process, so going for a Processing/JS form of teaching actually makes this something my brain can at least attempt to process.

I'm not a complete beginner when it comes to the concepts of coding — I know what a function is, I can understand for loops, and stuff like that —, but I never really tried to code. And how easy this thing proves you know shit even if you think you do. They start by just telling you to create a simple snowman, without ever needing to call more than the ellipsis function and define it's parameters, and you think "but that's too easy".

And then you realize that you actually need think about how ellipsis function works to draw it, what are the coordinates you need to use so it appears in you canvas, and make the math on where the middle is. Let me be clear about something: I can't do 2+2 without a calculator. My bachelors degree is in film studies.

They actually tell you where to draw things, and give hints on the parameters for some stuff, and I thought "nah, I'm gonna just think it through, this is definitely mean for kids".

The actual snowman itself turned out alright, doing that...

Like, it looks okay. I had to go through the parameters a couple of times to get it right, but it worked and my brain wasn't even screaming.

But then they tell you to put arms in the thing.

It turns out that the line function works in a completely different way than the other things you've learn so far. You have to tell it the coordinates of both starting and end point.

I have slight discalculia. Math in high school was a legitimate trauma, and this thing just threw me back into the hell that was analytical geometry. Fun.

From now on I'm just gonna take the hints and use the numbers they tell you to use and spare my brain from the pain of having to understand where things go. I'll just trust their method.

Between trying to remember the size of the canvas, where the coordinates start, and how each function take it's parameters and how they work, this is a lot of information. It will probably get easier as time goes by and I get used to all of it, but for a first time, it's a lot of things to remember.