Programming thread

[HexFag]

I get the sense this is the sort of thing the root of the problem is. Mastering time complexity is rapidly becoming a lost art; the compiler can't save you if you're just a retard with a basic grasp of loops. Some of my buddies were complaining recently about the slow dogshit their students today write compared to just a few years ago. Beginner programmers never test their things at scale.

When I was taking a class on Java, I learned that lesson the hard way when I needed to find a fast DFS through a massive CSV graph between two nodes of arbitrary distance. For the life of me I couldn't figure out a solution that didn't take literal fucking hours until I realized that I was adding values to my graph in the dumbest way possible. Ever since then, I've always been super anal about optimizing loops so that I wouldn't spend a literal eternity testing to see if something worked.

It also helps that I'm not using Java which is a piece of shit language, but the error was completely on my end.

 

[King Of The Kiwis]

I'm trying to learn react and I'm finding it kind of overwhelming. There's so many versions of it I don't know which one to use. Some tutorials use "Vite" and others use "npx create-react-app".
I also don't know what resources/videos are outdated and which ones are up to date. Can someone give me a React roadmap? I'd really appreciate it.

 

[The Ugly One]

I get the sense this is the sort of thing the root of the problem is. Mastering time complexity is rapidly becoming a lost art; the compiler can't save you if you're just a retard with a basic grasp of loops. Some of my buddies were complaining recently about the slow dogshit their students today write compared to just a few years ago. Beginner programmers never test their things at scale.



This hurts me to my core. Part of a thesis I did involved doing heavily optimized GPU linalg. If I did this shit, I never would have completed the main task in the next decade.

It gets better. The super-genius who replaced linked lists with arrays and got promoted to head of R&D, then fired everyone who knew things? He completely failed to figure out how to use OpenMP...like...at all. Couldn't get it to compile. Consequence? OpenMP was banned at the company, and mentioning it was basically asking for him to find an excuse to fire you.

I've seen things like this often enough to realize that any "genius" is really just a sociopath who steals credit, exaggerates his accomplishments, lies about his failures, and strategically runs people out of the company who threaten to outshine him.

 

[Creative Username]

It gets better. The super-genius who replaced linked lists with arrays and got promoted to head of R&D, then fired everyone who knew things? He completely failed to figure out how to use OpenMP...like...at all. Couldn't get it to compile. Consequence? OpenMP was banned at the company, and mentioning it was basically asking for him to find an excuse to fire you.

I've never used OMP but it looks like a standard header-and-object-based library with a bit of additional compiler pragma magic to make it easy to use. If I had to guess why it didn't compile I'd guess it was an omitted linker flag, perhaps? That's a common rookie mistake when using any library that supports threading.

I've seen things like this often enough to realize that any "genius" is really just a sociopath who steals credit, exaggerates his accomplishments, lies about his failures, and strategically runs people out of the company who threaten to outshine him.

Competent people do exist in this world, they just never worked there, judging from the outstanding quality of the code and the propensity of the resident team leader to kick out people that might be better at their job than him.

 

[DavidS877]

Competent people do exist in this world, they just never worked there, judging from the outstanding quality of the code and the propensity of the resident team leader to kick out people that might be better at their job than him.

Me:

 

[y a t s]

When I was taking a class on Java, I learned that lesson the hard way when I needed to find a fast DFS through a massive CSV graph between two nodes of arbitrary distance. For the life of me I couldn't figure out a solution that didn't take literal fucking hours until I realized that I was adding values to my graph in the dumbest way possible. Ever since then, I've always been super anal about optimizing loops so that I wouldn't spend a literal eternity testing to see if something worked.

It also helps that I'm not using Java which is a piece of shit language, but the error was completely on my end.

Having the insight to know which order to do certain operations, how different ones could be combined into a single loop for efficiency, and the like is the key thing. It is something gained through a bunch of practice, which a lot of individuals in paid positions have not put in the hours for. This is my main issue with the Pajeet programmers; it is fine to be a beginner, but it is not fine to also hold a paid position in a company to make production quality software if you have no clue what the fuck you're doing.

A good simple example is a recent commit I did for the SNEED Overlay to save superchats in the server state instead of the browser JS client in case a tab goes tits up. All messages are the same under the hood, so it was a matter of checking if a dollar value is present and saving them in a simple vector. By moving the Message object at the end of the handler function, it makes it so that I don't have to do a wasteful copy to have ownership in that vector.

I've seen things like this often enough to realize that any "genius" is really just a sociopath who steals credit, exaggerates his accomplishments, lies about his failures, and strategically runs people out of the company who threaten to outshine him.

I think it's mostly that genuinely smart people typically have a degree of intellectual humility and are constantly aware of the fact they may be incorrect about certain things. One of the most important things to learn about society is that people will follow anyone who says anything—regardless of validity or veracity—with utmost confidence and authority.

 

[Creative Username]

Having the insight to know which order to do certain operations, how different ones could be combined into a single loop for efficiency, and the like is the key thing. It is something gained through a bunch of practice, which a lot of individuals in paid positions have not put in the hours for. This is my main issue with the Pajeet programmers; it is fine to be a beginner, but it is not fine to also hold a paid position in a company to make production quality software if you have no clue what the fuck you're doing.

This comes with practice and a conscious commitment to doing things in an algorithmically efficient way. There is a general problem with people paying attention to the quality of their work. Even the most skilled programmer will write ludicrously slow code if he doesn't give a fuck about what he's writing.

I think it's mostly that genuinely smart people typically have a degree of intellectual humility and are constantly aware of the fact they may be incorrect about certain things. One of the most important things to learn about society is that people will follow anyone who says anything—regardless of validity or veracity—with utmost confidence and authority.

Truly smart people always act like they have no clue what the fuck they're talking about, because they are smart and recognize that the world is too complex for them to understand. See the Terry Davis bird quote. In turn, this knowledge of a lack of knowledge makes them more prone to learning stuff.

I guess the next layer of smartness is knowing when to lie and admit certainty so you can efficiently lead the masses.

 

[Safir]

I get the sense this is the sort of thing the root of the problem is. Mastering time complexity is rapidly becoming a lost art; the compiler can't save you if you're just a retard with a basic grasp of loops. Some of my buddies were complaining recently about the slow dogshit their students today write compared to just a few years ago. Beginner programmers never test their things at scale.

This is horrifying. I remember failing a test in a vocational training class because I wrote code which I thought was faster and "safer" checking for key presence in a map ("do not rewrite if exists"). Same O but slow enough to fail on speed.

 

[Whatevermancer]

I'm trying to learn react and I'm finding it kind of overwhelming. There's so many versions of it I don't know which one to use. Some tutorials use "Vite" and others use "npx create-react-app".
I also don't know what resources/videos are outdated and which ones are up to date. Can someone give me a React roadmap? I'd really appreciate it.

I think Vite is the cool new thing and I've heard that it works quite well, so I'd suggest setting up a project with it and then following the official react tutorial on the homepage.

 

[Nitro!]

I'm trying to learn react and I'm finding it kind of overwhelming. There's so many versions of it I don't know which one to use. Some tutorials use "Vite" and others use "npx create-react-app".
I also don't know what resources/videos are outdated and which ones are up to date. Can someone give me a React roadmap? I'd really appreciate it.

Weren't you saying a few pages ago you're just starting out with programming? Don't poison your mind with JavaScript, friend.

 

[y a t s]

Even the most skilled programmer will write ludicrously slow code if he doesn't give a fuck about what he's writing.

If I give enough of a fuck about the task at hand, I will fixate on it to the degree I forget to eat or sleep kinda like that flip-top box guy with his car. Unsurprisingly, this is when I did my best work/studying. It used to be really hard for me to focus on things I didn't give a single ounce of a fuck about when I was young, which was misattributed to ADHD. In reality, it's simply a lack of discipline.

I guess the next layer of smartness is knowing when to lie and admit certainty so you can efficiently lead the masses.

Benevolent social engineering. I like it. People who are the most successful are usually the ones who best know the rules/laws (of physics, society, what have you) and how to use them effectively to get ahead.

This is horrifying. I remember failing a test in a vocational training class because I wrote code which I thought was faster and "safer" checking for key presence in a map ("do not rewrite if exists"). Same O but slow enough to fail on speed.

It's funny how all of those little checks add up. Shit's going to get really bad as the average file or dataset size sharply increases over time. AI models immediately come to mind.

Weren't you saying a few pages ago you're just starting out with programming? Don't poison your mind with JavaScript, friend.

lol my first was C++ when I was in elementary school. I didn't get into any crazy template stuff; what I did was closer to C but not quite pure C (memory management, pointers, etc.). I think this had a big influence on me becoming the low-level lunatic I am today.
I technically did a small amount of HTML before that, but I don't count markup.

 

[Flatline]

Can someone give me a React roadmap? I'd really appreciate it.

Start with the React documentation.

 

[SIMIΔN]

This is horrifying. I remember failing a test in a vocational training class because I wrote code which I thought was faster and "safer" checking for key presence in a map ("do not rewrite if exists"). Same O but slow enough to fail on speed.

That's too harsh. Computers are so fast these days that linear increases in complexity don't really matter.

Anyway, I found a thought provoking video on how AI will change our industry.

 

[The Ugly One]

I've never used OMP but it looks like a standard header-and-object-based library with a bit of additional compiler pragma magic to make it easy to use. If I had to guess why it didn't compile I'd guess it was an omitted linker flag, perhaps? That's a common rookie mistake when using any library that supports threading.

I figured out how to do it on my own. It's not hard at all. This guy was just really not that smart.

Competent people do exist in this world, they just never worked there, judging from the outstanding quality of the code and the propensity of the resident team leader to kick out people that might be better at their job than him.

Sure, there are lots of competent people out there. What doesn't actually exist is the super-genius whose once-in-a-generation brilliance is the lynchpin of the company's success, who is solely responsible for throwing his enormous brain at nigh-unsolvable problems and doing magic. Genuine smart people know they don't know everything and are capable of learning from others, and they actively seek out even smarter people to work with. But by being actually smart and competent, they don't create the godbrained myth about themselves.

Every time I've seen management speak of some super-programmer in damn near reverential tones, the guy's turned out to be a fraud.

 

[Blamo2000]

Is programming still worth learning?

 

[y a t s]

Is programming still worth learning?

I think so, but only if you are genuinely interested or have some use for it. Learning at least some basic shell scripting can save you a lot of time and energy.

Keep in mind, there will need to be people who inspect the code output of AI models to make sure there are no issues or vulnerabilities in it for the foreseeable future.

 

[Blamo2000]

I think so, but only if you are genuinely interested or have some use for it. Learning at least some basic shell scripting can save you a lot of time and energy.

Keep in mind, there will need to be people who inspect the code output of AI models to make sure there are no issues or vulnerabilities in it for the foreseeable future.

It's a fascinating subject, and I think I could organize my work better if I knew how to compile and visualize data better.

I have a dumb question. I don't want to power level so I won't give software specifics.
Would it be possible to make a Python program that actually puts in data to the browser then "pushes" specific buttons. I am working with some browser based programs, but it's manual like hell. Imagine having hundreds of accounts and you need to pull their ledger manually one by one in Excel format.

 

[Marvin]

I have a dumb question. I don't want to power level so I won't give software specifics.
Would it be possible to make a Python program that actually puts in data to the browser then "pushes" specific buttons. I am working with some browser based programs, but it's manual like hell. Imagine having hundreds of accounts and you need to pull their ledger manually one by one in Excel format.

It's not ideal, but yes.

https://selenium-python.readthedocs.io/

An API specifically designed for automation is preferable over controlling web browsers, but that isn't always an option.

 

[Blamo2000]

It's not ideal, but yes.

https://selenium-python.readthedocs.io/

An API specifically designed for automation is preferable.

Then I guess, I will hate boring stuff to automate. Of course gotta learn this stuff first.

 

[Creative Username]

It's a fascinating subject, and I think I could organize my work better if I knew how to compile and visualize data better.

Python has some amazing data visualization libraries.

I have a dumb question. I don't want to power level so I won't give software specifics.
Would it be possible to make a Python program that actually puts in data to the browser then "pushes" specific buttons. I am working with some browser based programs, but it's manual like hell. Imagine having hundreds of accounts and you need to pull their ledger manually one by one in Excel format.

You could do that with a browser extension written in JavaScript. You can also use what @Marvin posted, if you must use Python.