After the (temporary) defederation announcement of earlier i checked the Lemmy repo to see if there was already a ticket on the federation limiting option like Mastodon’s that people mentioned Lemmy doesn’t yet have. Not only i didn’t find it, i also saw that there’s about 200+ open tickets of variable importance. Also saw that it’s maintained mostly by the two main devs, the difference in commits between them and even the next contributors is vast. This is normal and in other circumstances it’d grow organically, but considering the huge influx of users lately, which will likely take months to slow down, they just don’t have the same time to invest on this, and many things risk being neglected. I’m a sysadmin, haven’t coded anything big in at least a decade and a half beyond small helper scripts in Bash or Python, and haven’t ever touched Rust, so can’t help there, but maybe some of you Rust aficionados can give some time to help essentially all of Lemmy. The same can be said of Kbin of course, although that’s PHP, and there is exacerbated by it being just the single dev.
May I ask you more about your experience attempting to learn? I don’t know if it’s because I can program but eight years seems a long time. In return, I’d be happy to offer you a few pieces for advice that could help?
That’s so kind of you!
I started by aiming for front-end web dev. I learned HTML & CSS (I know, we’re not PROGRAMMING yet). At the time, that’s all I was hearing it’d take to get into the role. Then it was “you should probably know some Javascript,” and I wasn’t ready for how big of a jump that was. By the time I started understanding it, it became “learn jQuery,” which I learned and used for a couple of small websites, then came the libraries…
“AngularJS is the future” well now I need to learn Git, compiling, CMD…
ReactJS starts becoming a thing and I say “seriously? I’ve learned enough of these things, quit moving the goalpost, React isn’t going to stick around”
Yeah…it definitely stuck around…but as an Open Source nerd, I got super excited by VueJS and started learning that. No jobs in that apparently, aaaand I no longer want to do web dev, especially since I never reached the point of enjoying coding, it was always a means to an end.
So there were two major issues for me: \
var person = { userInput } var num; function findNumberOfLetters (person) { num = length(person); } function response(person, num) { findNumberOfLetters(person); console.log("Hello " + person + "! Did you know that your name has " + num + " letters in it? Numbers are rad!"); }I’m sure I did things wrong, but again, this is just for the sake of example. So, I write something like this thinking that it’s nicely structured and easy to read, and inevitably won’t work. I pass this to a friend, and the answer seems to always be a less structured, more nested code. So for this example, something like \
(function response(userInput) { console.log("Hello " + userInput + "! Did you know that your name has " + length(userInput) + " letters in it? Numbers are rad!"); })Obviously their answer is shorter and this isn’t exactly a complicated program, but for some reason, making the thing that provides a response to the user to do any of the logic feels wrong and messy to me. It’s a really hard thing to explain, I hope this makes some amount of sense, but I just process things very differently than code does, and it just ends up really incompatible. I’ll beat my head for weeks over-complicating something because I want it to “be clean” only for someone I know to come up with something that actually works within seconds.
This isn’t to compare my skills against them, it’s to say that I’m thinking about it wrong, I’m organizing it wrong.
That said, knowing how code functions has helped me to know enough to be dangerous and apply it in other ways, such as building Azure Logic Apps to manage ticket intake, or building alert monitoring. So it’s come in very handy in its own way, I just don’t have to competency to actually make any contributions to Open Source projects, especially since I don’t write in two of the coolest languages that I wish I could learn: Python and Rust.
No pity party here, I love what I do and I don’t intend to change careers to coding, but I do end up feeling helpless in the face of instances like this where I wish so badly to contribute to a project that I care so strongly about and want to see thrive. I know there’s other ways to contribute, such as providing graphics, UI, documentation, financial, hosting, etc. but coding always seems to be the most in need to keep up with demand, and with more and more projects coming out all the time, the more programmers are in need to see them through.
Going front-to-back
The entirety of your first paragraph is a humongous red flag. Your rationale makes sense, and it appears at each stage of the process you made logical decisions but the end result is disaster. You are evidently a knowledgeable and intelligent person, so I’ll ask you now to consider how old the underlying technologies on the web are, and how the web looked when they were written. Web frameworks exist to bridge the gap between languages and methods that are popular in the present day and the arcane mysteries of the world wide web. They bridge this gap with magic.
Now, this magic is a wonderful thing. It enables developers to make applications for the web in a way they’re familiar with and saves them having to think about the ‘web’ part which is:
However, if you’re trying to ‘learn programming’, starting with web dev is backwards in my opinion. You either choose to use a framework, and then you’re not really learning programming: you’re learning the framework. I always find this unrewarding, as it feels like the framework is doing all of the work and robbing me of the satisfaction. Furthermore, when you’re in the ‘beginner mindset’, you want to know how things work, you want to understand, you want to learn. Web frameworks are powerful when you focus on the ‘what’ and less on the ‘how’. Experienced developers have an understanding of ‘OK, this is what the framework abstracts for me, I’ll let it do its job and I’ll do mine’ and ‘Oh, I am not sure about this and it’s important I learn’. This is obviously a skill only gained by experience.
Furthermore, as you touched on, web frameworks go in-and-out of fashion and can have radically different approaches. With professional experience, you can apply your knowledge of one framework to learning another. As a beginner, I do not believe you have a broad enough understanding to be able to make connections between frameworks. This leads to your feeling of always being on the back-foot.
My recommendation for anyone who wants to learn programming in-and-of-itself, rather than to a specific end, is to start low-level. C was my traditional recommendation, though I may be altering this to Rust. C will never let you forget that you are programming a computer. This might sound ridiculous but it’s remarkably easy to lose track of and once you do, everything (in my head, at least) can get a bit esoteric and conceptual, and I find the ‘grounding’ deepens my sense of satisfaction working with computers.
My diagnosis as it stands is that you attempted to work from ‘abstract’ (web frameworks) to ‘concrete’ (learning how to program) and I’d always advise the reverse. Use C to gain a clear appreciation for how a programming language is altering 1s and 0s on your machine. Then build up with other abstractions, languages, and, in due time, frameworks.
Thinking differently
It sounds to me that, in your experience, you made some correct assumptions and some very incorrect assumptions. Correctly, you identified that structure, ease-of-understanding, and writing your code in a style that you’re comfortable with is optimal. Incorrectly, you assume:
Now, these aren’t merely incorrect assumptions, they’re incredibly incorrect. In the professional sphere, those who write code that is well-structured, easy-to-reason about, and is well documented are sparse but they are often highly valued (in good teams, at least). I was always taught, and still vehemently believe, that clarity and readability are absolutely the first priority. Speed of execution is a secondary concern. Terseness is not even a concern of mine at all.
Good code is easily understood by computers, as well as human beings. You are the first human being who has to understand your code. You should always write it in the way that is most easy for you to understand. It is a darn sight easier to optimise neat code, than it is to neaten optimised code.
I believe you’ve come to these conclusions because of your experience learning web development. I find web frameworks and JavaScript as a language to have syntax that is odd and inflexible. The pseudo-code you provided was in JavaScript and is reminiscent of many of the traits that mean that even experienced developers such as myself find it difficult to parse. There’s a reason callback hell is a meme. If you get the opportunity to use some other languages, I believe you’ll see there’s more flexibility in expressing yourself than maybe you’ve had the opportunity thus far.
Furthermore, I am reticent to comment on what is valued in the JS community, as I have never written anything of note with the language but looking at the two code snippets you provided, the first snippet smacks more of a ‘professional’ developer. It’s more modular, finely-grained, expressive.
Finally, if you’re still clinging onto the notion that coding demands a ‘particular’ way of thinking, I encourage you to look-up internet rants from people who know a lot about Java, Python, C++ etc. bashing Haskell. Haskell is a beautiful language that forces its developers to write in a style known as ‘functional programming’. I won’t go into depth about the differences and why functional programming is amazing but let’s put it this way:
In Haskell, you cannot do either of these things and it’s for the better.
Conclusions
I’m glad to hear you’re happy in your work but disappointed that you were unable to find a groove with programming. From what I’ve read, you absolutely seem to be the kind of thoughtful individual who should enjoy it.
I’d encourage you to start learning Rust. I am learning Rust at the moment myself. Pay close attention to references and memory management. When your material discusses memory concepts, such as the heap and the stack, ensure you develop an intuitive understanding of how they work. Don’t just view this as an important part of your development (one many developers neglect!) but as a window into what is happening on your machine. If your brain is anything like mine (and I think it might be), it’ll further your satisfaction.
Avoid Python. Python is tremendous is you want to use a tool that mandates Python. It will hinder your learning and force you into bad habits if you learn it early. Furthermore, when your project gets to any level of complexity, it’s going to be far more pain than its worth. I understand why Python is so popular in some non-programming domains but it has no appeal to me as anything more than a scripting language with boundless 3rd party libraries.
Look at functional programming, if you’re curious. This free book on Haskell is popular for a reason. Functional programming ideas are slowly making their way into all mainstream languages, for a very good reason. They force you to think differently, and it might be different in a way you prefer.
Thank you so much. This is a terrific insight. A little later I’m going to sit down at my PC and type something up I think will help.
I’d ask you as a fellow ADHDer ✊ if I forget to please remind me to write up my response 😂
EDIT: Response now posted here
Likewise, you gave an amazing response and I want to dedicate some time to properly reading and responding to it, so I’ll be responding when I have a good opportunity to commit :) Thank you so much for such a thoughtful and thorough response, I really look forward to going over it
I’m a programmer of 25+ years. Everything written above is spot on. I too started with C and I still love that thing like my first born. It is so immensely satisfying writing something in C that a) works b) doesn’t leak memory and c) passes all your unit tests. Nothing else compares.
I too looked at React and hated it with a passion. Then I saw VueJS and kind of liked it. Then I saw Nuxt and now I’ve gone all in on NuxtJS. It is so simple and well-thought out compared to the shambles that is React. It’s very satisfying to use. Rust is next for me.
I’ve learnt well over 10 languages over the years. Some well, some well enough. Learning a language is bit like reading a book. If you’re a third of the way in and it’s doing nothing for you, don’t waste your time. Grab another one off the shelf and try that. Don’t put pressure on yourself - it should be enjoyable, not stressful. Just chip away at it bit by bit and enjoy the little discoveries.
Don’t worry so much about your coding style. From the examples you gave, yours is much easier to follow than the second one. And, you know what? Most ‘senior programmers’ I’ve worked with have been bad coders. The bar is not as high as it may appear.
Sounds like you’re currently a systems guy with a bit of programming skills. That’s an awesome combination to have, and mirrors my own all those years ago. The best bit is your have the freedom to learn programming as a hobby, without the pressure. Enjoy the process. Watch some videos by Sebastian Lague on Youtube, they’re magical.
Best of luck with your programming journey. It is an immensely enjoyable hobby, and ridiculously useful skill to have.
Thank you for taking the time to read my response. I really appreciate having the approval of someone with decades of experience (which, I very much don’t). Out of curiosity, when you started programming would have been the early days of Java, C++ etc. and the start of the ‘OOP revolution’. Can you recall why you started with C, when OOP was very much en vogue?
That’s a very good question. Java and OOP really were new and exciting things when I first entered the workforce. But, I first started programming long before Uni and joining the workforce. I first started programming on one of these: https://en.wikipedia.org/wiki/Intellivision using an ‘Intelliputer’. Then I learnt Logo on an Apple IIe. Then GW-BASIC on MSDOS, then PASCAL, then C. Then I stuck with C for a long time, including most of Uni.
So, by the time Java and OOP came along, I already had 10+ years of procedural programming behind me. That actually made it really difficult for me to to learn OOP. It was some weird paradigm that didn’t align with anything I knew. So I gave Java a mostly wide berth for as long as I could and focused instead on Perl, then PHP.
Both of those ended up getting OOP abstractions crammed into them, so eventually I did end up learning OOP, but it was more through osmosis. By then Java had largely come and gone. As much as it saddens me, Perl too largely died. That left me with PHP which, to be fair, was becoming quite a nice language by then.
By now I was a devout OOP developer and really enjoyed the OOP changes to PHP. Python, being object based, was the next logical choice. Javascript was still a pile of poo. Then Javascript got OOP abstractions too and suddenly was all the rage. Then they started putting the frontend on the backend (Node) and I was truly lost at sea. To me it it was twice as much work to build something (API + frontend) compared to the old server side rendering. It seemed way to much effort just to avoid a page load.
Nevertheless, I tried React and I hated it, it made no sense - though Javascript was looking a lot nicer by then. So I held off a little longer on adding ‘frontend dev’ to my resume, then NuxtJS came along and here we are.
So, long story short, I’ve been doing this forever. I technically review team code all day, every day. I know good code, I know bad code and have seen shit that still gives me nightmares.
The funny thing is, I have written dozens of cool tools that I think are pretty damn clever, yet I have never open sourced anything. My reason is the same as yours: Because I still don’t think any of it is good enough.
Here is a thing I found helpful in my >10 years of programming career: break your problem from top to down (aka Top-Down programming).
Consider this problem: I want to send a post card to my friend.
Now let’s break down this problem into several steps:
Next, each of those problems can be broken down even further.
Obtain a post card:
Now each of those steps can again be further broken down into even more sub steps:
Visit a store that sells a post card:
You see where I’m going? Break your problem down onto actionable pieces you can solve in the programming language of your choosing. How deep you go depends on what programming language to use. If you’re using a High Level programming language such as Python, you usually don’t need to break down the problem too deep in order to solve it. But if you’re using a Low Level programming language (e.g. assembly), you’re going to need to break down your problem very deep into actionable pieces that can be solved in assembly.
Hope that can help you.