Saturday brings #4x4
Besides my "Friday brings Python," I decided to write more about things I like. Did you know that AWD and 4x4 isn't the same thing?
Saturday brings #4x4
Besides my "Friday brings Python," I decided to write more about things I like. Did you know that AWD and 4x4 isn't the same thing?
Friday brings #Python
It's time to post about ELSE for loops. Yeah, that's right. Python allows you to pair ELSE with FOR or WHILE in a way that the ELSE block executes only if the loop was finished naturally without interruption by BREAK.
Never stop learning, especially nowadays, when LLMs can automate the boring stuff for you.
Spending time in nature is an awesome way to recharge.
I love coding, but I should accept that using LLMs for boring stuff is a good thing.
Nowadays, big players subsidize LLM-coding. In some areas, that makes junior or even more professional devs useless. What would the real prices be then? I mean, wouldn't it be cheaper to have one person plus one AI instead of three AIs with bad outcomes?
Coffee is an interesting thing. It used to produce code. Now, as AI replaces humans, we’re not converting coffee into code anymore.
I miss the days when I was young and had so much time.
For instance, it was building an HTTP web server in plain C and working with unix sockets, just for the sake of learning and fun.
LLMs are just token prediction. Stop seeking conciseness there.
Friday brings #Python
Today I have something special for you. If you’ve ever used generators in Python, you’re probably familiar with the `yield` keyword. But what about `yield from`?
Take a look at the image: it lets the current generator delegate control to another generator until it’s finished.
Testing bullshit theories? That’s a pretty nice benchmark for LLMs.
They said math isn't useful. Also, AI is made of:
Does starting with C/C++ compared to starting with Python make you a better programmer?
Why do people prefer to honk their horn when somebody in front doesn’t move as fast as they’d like?
Fishing is a nice way to rest your mind.
Look at this beauty!
Friday brings #Python
Today, let’s explore one item from the itertools package: cycle. This function lets you literally cycle through a list (or other iterable), which is very handy when you need rotate.
Imagine you need to build a turn-based game. Check the example.
There is a rubric called “Friday brings Python,” where I share Python tips. I’m considering doing something similar—maybe “4x4 Saturdays”—to share thoughts about cars, AWD, 4x4, etc.
GPT‑5.3‑Codex or Claude Sonnet 4.6?
There is an interesting project that could make inference quite cheap. They “hardcoded” the model into a hardware chip.
Imagine being able to run a modern model very fast on your own hardware.
Agentic coding is fine when you understand what you’re doing—and what the AI produces.
Did you know that LLMs are very good when it comes to copying existing content?
Aren't you surprised they were able to make a C compiler then?
Friday brings #Python
Imagine you want to write code to roll a 20-sided die until it gives you a critical hit: 20 points!
You have to use a while loop, right? Right?
When it comes to comparing owning a house vs. renting, there’s a lot of debate. Recently, I’ve shifted my opinion on it a little.
A mortgage is risky, but renting a house has a 0% chance of making you its owner.
OpenAI is directing resources away from long-term research and toward monetizing ChatGPT.
The only way AI will replace good software engineers is with a new generation of LLMs. Without more resources being thrown at that, there won’t be any next-gen LLMs.
I feel safe now.
Elon fired software engineers, but they said X works very well.
Usually terms “gross salary” or "super gross salary" are ignored.
Why should we care if the employer pays those taxes?
Technically, the employer does. But in fact, you receive less money if taxes go up. Isn’t it important to know how much the employer spends?