Yeah like, all the information you’d need for the moving parts of this bad solution come down to just one value, which is the length of the string. That’s the number of place values you need to consider, and also the highest value digit to consider. It sounds kinda like someone whose programming history is a bunch of functional but imperfect Project Euler solutions. I have trouble imagining that anyone who wasn’t based out of python would seriously consider this as a viable alternative to literally anything. It sounds like an astounding use of python’s extreme flexibility with variables and large selection of prewritten features.
Ugh, R. Well, there are definitely worse things that still count as coding, so I guess R has to as well.
Gang. How can I teach myself to be an employable programmer without getting incredibly discouraged? I've lost half my income from Covid, and I'd really like to gain some employable skills. I'm okay with HTML/CSS and I'm trying to learn PHP. And I did some Python in college. But at the moment I'm stuck staring at project suggestions (e.g. "Predict Movie Success through Data Mining") going "what? What? how? And why?"
Wat. Why don’t they just ask you to solve for fashion while you’re at it. That sounds like common sense encountering the dangerous edges of the Big Data explosion. The edges that have been up to some shit like trying to reinvent phrenology (yes, really) and other things where people decide that maybe they can work with the horribly flawed idea if they throw a lot of data at it. I don’t think it’s ethically wrong to try to use data mining to predict movie success if you’re not trying to convince anyone to buy into it, but I do think it’s stupid. If you’re particularly mathematically inclined, maybe Project Euler? If you’re not, that leaves every other approach to programming, and I have no idea what resource covers all of that. It also depends on which language you’re using for what is a reasonable learning project. But if you are looking at projects like that and losing interest, I’d say you have good instincts.
Rust is everything I ever wanted in a strongly-typed language and it's absolutely kicking my ass and I hate it I hate it I hate it Rust is the best. Someday my code will compile and I'll be able to move on to giving it actual runtime behavior, as opposed to it just being a long list of unused type definitions like it is now.
A variant of the Y2k problem: Berkshire-Hathaway Class A shares are trading at more than $421,000, and the market looks good for them. Nasdaq, and indeed a lot of the big stock exchanges, store stock prices as 32 bit unsigned integers, interpreted as .01% of a dollar. 232 is 4,294,967,296, which means that the highest price that Nasdaq can comfortably represent is $429,496.7295 So Nasdaq temporarily halted trading of BH Class A shares until they can upgrade their system.
There's programming, there's metaprogramming, and then there's programming via making an AI do it: Fun and Dystopia With AI-Based Code Generation Using GPT-J-6B
I'm going to start referring to "debugging" as "delousing". See how long it takes for my coworkers to murder me.
Cross-posting from the math thread: implementing machine learning (and other things) using only linear functions (subject to IEEE floating-point conventions):
I actually kind of regret getting so much glee out of the above video because I had to actually think about error propagation from IEEE floating point conventions today and it did not make me happy.
